Transcript of Episode 43: Erich Jarvis

Grant: [00:00:00] Welcome to The Bioinformatics CRO Podcast. I’m Grant Belgard, and joining me today is Erich Jarvis. Eric is Professor at Rockefeller University and investigator at HHMI. Thanks for joining us, Erich.

Erich: [00:00:11] You’re welcome.

Grant: [00:00:12] So can you tell us about what your lab does?

Erich: [00:00:15] My lab studies the brain basically, but we study more specifically brain regions that control our ability to imitate sounds like that I’m doing now, producing the learned speech. And we do so in non-human animals that have that ability, like parrots and songbirds. These are rare few species in the world that can do this that are shared with us, even though they’re not closely related to us. So this is why parrots can imitate us where a dog can’t. We also have a genome lab that is generating lots of high quality genome data, not only for our neuroscience projects, but for the scientific community broadly, ranging anything from neuroscience to conservation.

Grant: [00:00:53] How do you think the genome effort can inform the neuroscience effort?

Erich: [00:00:57] Most traits are controlled by what’s in the genetic code of your genome and the genes in your genome. And so if we find this specialized trait like spoken language in humans that we don’t find in our closest relatives or like in parrots, we can compare the genome of humans in chimpanzees or parrots and falcons and ask the question, what differs there in their genetic code? And the more species you have with these traits, the higher the N number, which helps you statistically find the specific genetic markers that will develop a brain pathway that allows you to develop a neural circuit for, let’s say, speech or learning how to fly.

Grant: [00:01:38] And how similar are these circuits between humans and birds with vocalization?

Erich: [00:01:43] What’s surprising is that some of the similarities outweigh the differences–when I say we, it’s me and other people in the scientific community that study this not just my lab–but overall, we find that humans and parrots and songbirds have a specialized forebrain circuit in the parts of the brain you would call the cortex and the basal ganglia that you don’t find in any species that cannot imitate sounds or find it to a very rudimentary degree. So the fact that they are there is similar, right. And then they have some connections that are similar. Like the cortical regions in us humans and the equivalent in birds project down to the neurons in the midbrain that control the voice.

[00:02:30] And this connection is direct from the cortex where you don’t find those direct connections in species that can’t imitate. What’s remarkable is that we’re separated from birds by 300 million years of evolution. We have so many species in between that don’t have these circuits. Further, what’s remarkable is that the mammalian brain, including us humans, the cortex is a set of layered neurons like six layers, one on top of another. Whereas in birds, the neurons are clustered in the cortex. But yet, from the layered structured or clustered structure, you get a similar type of neural network that controls speech.

Grant: [00:03:08] So is this a case of convergent evolution?

Erich: [00:03:11] Yes, convergent evolution. It happened more than once in evolution, and it happened in similar way.

Grant: [00:03:16] Do you have any idea how many times this has happened?

Erich: [00:03:20] Yes. So far we think at least five times in mammals, us humans, whales and dolphins that together are called cetaceans, bats, elephants and then another marine mammal, pinnipeds. Those are seals and walruses and so forth. So it’s five out of like thirty two major orders of mammals. So you don’t find it in horses, for example. And amongst birds, we have parrots, hummingbirds and songbirds. So far no reptiles, no fish, no amphibians are known to have this trait.

Grant: [00:03:52] And is the kind of neural solution to the vocalization problem, for example, in mammals always the same?

Erich: [00:04:00] Yeah, it’s similar. So the similarity is that in all the species we’ve looked at so far, which is basically the three bird groups in humans compared to their closest relatives. I would call this spoken language pathway or the vocal learning pathway, I think they’re nearly equivalent that it is embedded inside of a motor pathway that controls learning how to move, like learning how to move the hands, learning how to move other parts of the body. And this motor learning pathway we can find in all vertebrate species that move right. So what we think happened is that the brain pathway for spoken language evolved out of the brain pathways controlling learning how to move and evolved by a brain pathway duplication. And so that’s why we think it’s similar in birds and humans because the ancestral brain regions out of which the speech pathways evolved were already there.

Grant: [00:04:54] How generalizable do you think the circuitry patterns are? Do you think that everything is built on this motor substrate and there are many, many paths to get to vocal learning and this is just the most common based on what was pre-existing?

Erich: [00:05:10] I actually think there are limited paths in which you can get to vocal learning. It’s like the evolution of wings, especially amongst vertebrates. Each time they evolved in bats, birds and ancient flying dinosaurs, the pterosaurs, they evolved on the upper arms, not one on the back or one around the tail and one by the feet and so forth. And the reason why is that there’s an environmental constraint. And that environmental constraint is the center of gravity. So if you’re going to fly in the sky, you need your wings to be near the center of gravity to fly more energetically. I think the same thing is happening to the brain. There’s limited ways you can evolve a circuit that’s going to control, learned sound production, especially through our vocal organs like the larynx.

[00:05:57] And I like to think of the surrounding motor pathway as the arms of the wings, right. And the vocal learning circuit as the wing structures themselves. But does that theme happen multiple times? I think my prediction is yes, no one really knows. But my prediction is that the motor learning circuits for different traits can emerge out of sort of canonical motor learning circuit and then become specialized. The specializations for speech in us humans and songbirds are of two forms. One is this is an advanced sensorimotor integrative behavior. We need to take sound coming through our ears and integrate it with movement of the jaw muscles and the laryngeal muscles and other things that control the sound production.

[00:06:46] That kind of sensory motor integration of auditory input motor output, I think requires its own special formulas–or algorithms in computer science terms–to work with each other. In fact, after vocal learning evolved it looks like only vocal learning species are the ones that can learn how to dance, synchronizing your body movements to rhythmic beats of sounds. And the reason why I think that is the case is because you need to synchronize rapid auditory input with motor output. So you need something special to have the auditory information talk to the motor information. And once that happened, it contaminated the surrounding motor circuits that control not just the larynx, but the hands, the feet.

[00:07:28] And so to have auditory input synchronize our body to rhythmic sounds was a side effect of spoken language. The second specialization I think that happened is the larynx has the fastest firing muscles in the body to produce sound. You need to move those muscles really, really quick to vibrate the air and produce sound or modulate the sounds. And we find that the neurons in the speech circuit for humans and in the vocal learning circuits of these birds are over enriched with molecules that control rapid-fire neurons that control neuroprotection so you don’t kill the neurons just by speaking. That’s another specialization you don’t find in the surrounding motor circuits.

Grant: [00:08:15] You brought up some pretty interesting things there. So you mentioned dancing is maybe almost a side effect. But when I hear dancing in the context of evolution, I think sexual selection, right. Do we know which came first?

Erich: [00:08:27] Yeah, we don’t. But there are a lot of theories out there of what cause vocal learning to evolve, and that includes spoken language and sexual selection is one of them. I can say that I tend to believe that because all the vocal learning species use their learned sounds for mate attraction or in case of non-humans for territorial defense. What we can do it for territorial defense as well, very few of them use it for more abstract semantic communication like we’re doing now to communicate ideas, to communicate concepts. Instead, they sing to attract mates. The more varied the songs or the more you steal sounds from the environment incorporated in songs like in mockingbirds, then more likely you’ll attract the opposite sex.

[00:09:16] Now, you would think semantic abstract communication, referential communication should be the first thing we use it for. What most people don’t realize—and even a lot of scientists–is that referential communication like using a sound to mean a bear, using a sound to mean this object over here, that’s already happening before even spoken language evolved. Like vervet monkeys have an innate repertoire of sounds that through cultural experience, they will learn to use for different predators or food and so forth. But the first thing vocal learners do with their learned sounds is to attract a mate. So that’s why I think it came first.

Grant: [00:09:56] In what species that have vocal learning, do you see rudimentary elements of culture? You’d certainly expect groups of dolphins and so on to learn from one another.

Erich: [00:10:07] Yeah. I’m going to answer the question you’re asking, but I think you might be asking a different question, right. Because all vocal learners culturally transmit their repertoires, whether they’re using it for semantic information or what we call effective information, like to attract mates. So it’s cultural once you have vocal learning. And by being cultural, you get different dialects, like we get different language and so forth. They further geographically you are separated. Now I think the question you’re asking on top of that is what species will culturally transmit information about their vocal repertoire that’s more informative like this sound means predator and so forth. And there are very few species that do that besides humans. Dolphins are thought to be one, corvid songbirds, which are basically crows, and Blue Jays and so forth, thought to be another. And the parrots, like African grey parrots.

Grant: [00:11:05] I guess another really interesting question that this raises is obviously comparative neuroanatomy and comparative genomics are enormously powerful tools to study this. But of course, a complementary approach is classic genetics or human genetics, right. I mean, looking at broken genes, seeing what they do, studying diseases and verbal defects and so on. How complementary have those been and have they pointed in the same direction, since work on FOXP2 and so on?

Erich: [00:11:35] Those two questions basically describe the approach we take in my research broadly. So yes, we use comparative genomics like a natural experiment. The more genomes you sequence out there with species with different traits, the higher probability you will find whatever genetic difference is responsible for those traits. So we sequence genomes and we compare genomes to find out if there’s convergence in all these different species with and without vocal learning. And if we find convergence, then we take those genes that have these convergent genetic changes. And what we’re doing now is taking the genes of a species that learns how to imitate sounds and putting it into the genome of a species that don’t like a mouse. And see if we can induce a change in the brain circuit to get us further along the path to becoming a vocal learner.

[00:12:34] And if we do, that proves that this gene is responsible for contributing to the trait and we can study what it does, its function and so forth. And if it doesn’t, we falsified our hypothesis. Another way of doing it is within one species like humans is you compare different people, and you find a family out there that has a speech disorder, speech deficit. They can do everything else fine, but they have difficulty in learning speech. And then you sequence their genome and you find something that’s different from them compared to all the people who can produce speech normally. And this is exactly what happened, more than a decade ago in the discovery of the FOXP2 gene. This is a gene, it’s a transcription factor, meaning that it regulates how much a gene product is made for other genes in the brain, that controls neural connectivity.

[00:13:30] And when this gene is mutated, it causes people to have difficulty learning how to produce speech. And we put that same mutation in mouse in collaboration with Simon Fischer in Germany. We found that even though these mice are producing mostly innate sounds, like humans they had difficulties switching to the more complex innate sounds that females like to hear in their courtship, right. We find this gene in songbirds and if you block its function in songbirds, just like in humans, it also prevents vocal learning from happening normally.

Grant: [00:14:15] So how do you think about translatability?

Erich: [00:14:18] For a long time, I was hoping that the work we were doing in songbirds–because I started out with songbirds–people would take those discoveries that we found and then try to translate it not only into understanding human knowledge, but to for human health. And I found that people were not doing that. So the first thing we did was find out if we could find convergent parallel changes in human genomes that we see in songbirds for vocal learning. And the answer is yes. Not everything is convergent for the genes but there is a lot of overlap in the genetics. And now what we’re trying to do is to see if we can induce a mouse to become more of a vocal learner species and study things like stuttering or autistic types of speech deficits and these FOXP2 mutations in mice. So that one day those can be translated to helping humans not only understand the disorder better, but to repair it.

Grant: [00:15:24] And how do you think about that in the context of developmental windows, right? Because by the time someone is diagnosed with a speech disorder usually that ship has sailed. But do you think that can be reopened in some way?

Erich: [00:15:37] Because you can have multiple people out there with the similar speech disorder all affecting the same genes. You don’t have to wait for somebody to grow up, to become an adult, to discover it. So by doing a population analysis as opposed to longitudinal analysis, you can get at discoveries quicker. Although waiting 13-14 years for a person to go through puberty and then find out whether they can produce normal speech and so forth is also necessary. That leads to that other part of the question the critical period of years. In all vocal learning species, there is a critical period or what we now call sensitive period, where it is easier to learn how to acquire speech early in life and later in life. And then once you pass puberty, it’s harder, like it’s harder to pick up a new language.

[00:16:24] There are certain set of genes that are turned up or turned down in the brain during those critical periods that close off the ability to learn as much as you can when you’re a child. And there are people who are now trying to switch those genes back on. So that allows you to learn, in this case, spoken language as well as you did when you were a child or at least getting closer to that. And I think that’s going to one day be possible, but it’s going to cause some problems also. And the problem that people don’t appreciate is that. Why can’t I learn as well as when I was younger? Why is it taking me so long to learn how to ride a bike as an adult then when I was a child?

[00:17:04] And the reason is that if your brain is in a very plastic stage where you can mold it and learn a lot quickly, you’re also going to forget quickly. And this is why, sometimes it’s hard to hold on to early childhood memories because you forgot a lot of it because there’s only so much capacity in the brain. If you’re going to learn, sometimes you’re going to erase. And yes, you can erase memories. So if you’re going to reopen the critical period, you better do it for a brief period of time. Learn what you can so you don’t erase a lot.

Grant: [00:17:35] This sounds like a premise for a good novel.

Erich: [00:17:38] That’s right.

Grant: [00:17:39] So what excites you the most about this field? What do you think is most promising, and what do you think are some of the biggest as yet unanswered questions?

Erich: [00:17:51] Yeah. What excites me most? Well, maybe I went into neuroscience because I was interested in something that was mysterious. The brain is one of the organs that we have the least knowledge about, but the biggest investment in or one of the biggest investments. I guess the biggest investment is in cancer broadly, but cancer affects the whole body. And so I’m talking about organ systems. Maybe the heart gets bigger investment. But I’m just fascinated by the fact that we have this kind of behavior or spoken language that allows us to culturally transmit knowledge from one generation to the next. It makes us humans the advanced species that we are. That’s what really fascinates me.

[00:18:32] Jumping many years later from me, starting this lab over 20 years ago. The biggest problem now is I would say mental health. It’s one of those things that’s a real mystery and is hard for us humans to figure out how to repair. It’s not as simple as stitching a wound and fixing it. And I think mental health is a bigger problem in humans than in other species. Not a lot of people think about mental health in other species, but think about your dogs who are home lonely and so forth, right. That can cause mental health disorders. Now people will listen to this, and they won’t want their dogs to be home alone. Get another dog to play with it, right. I think the problem is goes to another gene that’s actually involved or interacts with genes involved in language and spoken language circuits.

[00:19:22] In us humans, we have extra duplication of a gene called SRGAP2, spelled out as SLIT-ROBO GTPase. SLIT is a molecule that binds the receptor called ROBO. When they interact, they influence connections in the brain. Those two genes, SLIT and ROBO are turned up or down in certain brain regions that control speech that we think control the connection to the muscles that control speech. I’m sorry, I’m giving you all these molecular terms that the general audience might know, but I’m going to do it anyway. This GTPase modifies this SLIT-ROBO interaction to influence connections in the brain. It either dampens down its function or enhances its function. And so we humans have an extra copy of this GTPase molecule.

[00:20:16] And what that extra copy does is it inhibits the function of the normal gene. And by inhibiting the function of the normal gene, we slow down brain development in humans compared to other species. So our brain is developing at a slower pace and staying in a more juvenile state throughout adulthood compared to all other mammalian species or vertebrate species. And I think it’s leaving our brains in a more immature state, which then leads to more mental health disorders compared to other species. So some of these mental health disorders, are a consequence of having a more advanced civilization and our brains staying in a more juvenile-state so we can continue to learn throughout life.

Grant: [00:21:04] That’s a fascinating hypothesis. So I guess in that case, you might expect that there could be SRGAP knockouts, assuming its survivable out there walking around. Do you know if anyone’s kind of looked into what those phenotypes look like? Are there people whose brains basically mature much, much faster?

Erich: [00:21:22] That’s an interesting question. Because I have never thought of that, but it’s actually a doable question. And it makes me think, in humans there are either one or two extra copies of this SRGAP2 with people walking around. And it’s making me think, why have two extra copies in some people? And I’m thinking, maybe because one extra copy is not enough, you knock it out and then we become primitive human beings. The additional extra copy is like a safeguard in case one of them goes awry. I don’t know. There’s lots of genome sequencing being done on many people out there nowadays, and this question can be answered, theoretically. It might be difficult because what a lot of people don’t know in the genome world is a lot of the sequencing that’s being done on humans out there is being generated with what we call short reads.

[00:22:15] These are nucleotide base-pair sequences that are like 100 to 150 nucleotides long, whereas the genome is three billion, right. Whenever you have repetitive sequences like the SRGAP2 duplications, with short reads, it’s hard to figure out which copy is which. So you need long reads, long reads are more expensive, like from PacBio, Pacific Biosciences and Oxford Nanopore. And in the genome world when we produce high quality data, we’re using long reads. So they’re really figuring out this question. To answer your question, we’re going to need to sequence the genomes of a lot of people with long reads and then look to see if somebody is missing these extra copies of SRGAP2.

Grant: [00:22:57] That’s interesting. Yeah. I just had a quick look on gnomAD as well. And there are six observed putative loss of function SNV’s when there thirty-eight expected. And there are exactly zero homozygotes out there in all the genome sequencing databases that Nomad aggregates. So even the heterozygote knockouts are pretty rare.

Erich: [00:23:18] Interesting. And to think that this gene is in extra copies in humans. So you would think if we lose it, we will be like all mammals and would be OK, but probably and not.

Grant: [00:23:31] Maybe it’s especially important. So all this kind of leads us almost into our next segment. We’re talking a lot about dancing and its relevance for vocalization, and you were on the verge of going down the path of being a professional dancer, right. Can you tell us more about that?

Erich: [00:23:51] Yeah, that’s right. Actually a lot of my family were into the performing arts, and that’s the direction I was headed in. And a lot of them sang. I was an okay singer, but not as good as the rest of them. So what could I do? I started dancing in dance clubs and so forth as a teenager, back when they allowed teenagers to go to dance clubs. And I started winning dance contests, and I thought, oh, so I can dance. And I auditioned for a high school of performing arts here in New York City and got in. And was on my way to becoming a professional ballet dancer and jazz dancer. But at the end of high school, you know, I was really trying to make that career decision that many teenagers are trying to make. What are you going to do when you graduate high school?

[00:24:32] And I liked science as well. And my mother always said, do something that has an important impact on society. And that stuck in my head and I was choosing dance or science. And I thought as a scientist, I could have a bigger impact on society than I could as a dancer. And I think as a dancer, if you become a well-known dancer, you can have popular influence like anybody in the performing arts like actors and so forth. But as a scientist, I can make a direct impact. So that’s why I chose science. I went to Hunter College here in New York City. I got into a laboratory they were doing bacterial molecular genetics, studying genes that are involved in synthesizing proteins.

[00:25:18] And I found out there that being trained as a dancer trained me to become a scientist. They both require a lot of discipline. They both require creativity. Lots of failure before success. They’re not nine to five jobs. And so, so many things. Basically, I consider being a scientist is also being an artist.

Grant: [00:25:40] It’s fascinating. I never really thought of that before. What do you think doesn’t translate as well beyond the obvious?

Erich: [00:25:47] Being an academic scientist is like running like a small business. You have a lab, you have a people in your lab, you have the staff scientists, and you have the students and so forth. If you’ve got to raise money, your publications or your product. The more you publish, the more likely you’ll get money. So let me backtrack a little bit on that, right. It’s not as ruthless as the business world maybe, but politics and science, they seem to be two different worlds to me, especially at this time. You know, especially in the past four years before the current era where politics seemed to be anti-science.

Grant: [00:26:22] It’s interesting, right. You had a postdoc, who went on to be a pretty prominent politician in Puerto Rico.

Erich: [00:26:28] Yeah, that’s right. Ricky Rossello, my former postdoc, the governor of Puerto Rico. And I guess he’s a scientist, but I guess his politics didn’t mix with science that well, because he stepped down. I very much appreciated his time in the lab. He and I got along very well. I mean, he was a very creative, forward-thinking person. But I do encourage my students and postdocs and others not just to go into academia, but to go into other walks of life and including in politics. And even though I said they don’t mix that well, I do think we need more scientists in politics to help the world.

Grant: [00:27:08] And outside of academia, which is obvious. What paths do you think the PhD route is especially good training for?

Erich: [00:27:16] I think the PhD route is training you for the kind of jobs that require lots of problem solving. You know, let’s say city planning in the business world as well. I think it helps with problem solving because in science, you’re always challenged with a mystery, an unknown that you’ve got to figure out and solve. So maybe being a detective, you know, maybe being a forensic scientist, something like this.

Grant: [00:27:44] And are you glad you did it? I mean, if you had it to do over again, would you be a scientist? Would you be a life scientist? Would you study vocal circuits?

Erich: [00:27:52] Well, I guess if I had to do over again. There are two questions that for me are like: would I come back to the same field? And outside of science, what would I have chosen? And within science, I was always fascinated with the origins of the universe. So I was trying to choose between the origins of the universe and of the brain. So I might choose that or how life began or something like that. Outside of science, I was fascinated with history. So maybe origins of human civilization, origins of culture. So I might have chosen that or maybe I would’ve just stuck with dance. Those are the ones that come to mind. You know, I also considered becoming an astronaut.

[00:28:32] Well, you can’t just say that. You have to competitively apply for that. To me, it’s kind of connected to science, but you know what I mean. I could have been flying planes or something like that. I have no idea. But that’s something I was considering.

Grant: [00:28:46] So it sounds like pretty much big questions like fundamental questions as opposed to kind of applied questions.

Erich: [00:28:54] That’s right. So fundamental questions and that’s what excites me. I’m as excited about being a scientist as I was being a dancer. Because I feel like I’m getting to fundamental principles, I’m doing something that is fun, even though, you know, it’s not fun to try to raise money and get rejected from grants or get your papers rejected and so forth. But hey, you know what, in the interim, I’m having fun.

Grant: [00:29:18] So walk us through your kind of key decisions in your career. You know, in college, you decided to go the life sciences route and then you stayed in New York for your PhD, right?

Erich: [00:29:29] That’s right.

Grant: [00:29:30] And what made you decide neuroscience over, you know, cosmology or something?

Erich: [00:29:35] Yeah. Well, let me give you a little bit more context because actually, I started out wanting to be a magician when I was a young child. And I was emulating Houdini as a teenager, going down different parts in the New York City with my cousin Sean to be tied up and escape from chains and ropes and so forth. And we would figure out how to trick people into believing that something magic happened when it really didn’t happen. And so that kind of actually got me into science. And plus my father was interested in science. Because I started getting tired of trying to trick people. I really wanted to know how things really work. And then, jumping years later into my transition into neuroscience, it was really connected to dance. I felt with dance, the brain controls dancing. It’s something I can hold in my hands. It’s here on Earth. I don’t have to look up to the sky to try to figure it out.

[00:30:30] And I don’t know, it was something I felt more biologically grounded. It was a simple holistic way of thinking, and that’s why I chose neuroscience. But during my undergraduate years at Hunter, I was still kind of undecided in that trajectory. You know, is it going to be something in the biological sciences like neuroscience or is it going to be or is it the universe? So I double majored in biology and in math because if I went into physics and astrophysics, I knew I needed a strong mathematical background so I would have the choice by the time I graduated, which one I was going to do. The mathematics gave me a decent bioinformatics foundation for biology.

[00:31:06] And nowadays, you know, biology is so heavy with big data that that mathematical background is helpful. And then I was going to add one more thing onto that, which is I was toying around with the idea of should I go into activism, into politics? My mother said, do something has an important impact on society. And I did think about politics and so forth. And even within the sciences, me as a person who’s an underrepresented minority of African-American descent and mixed up with lots of other things, I thought about becoming more active in trying to change the culture. And I got asked to do this a lot, but I find that it’s actually like having two jobs.

[00:31:50] And so I figure I want to make a change in society by being a role model, by being an example that as opposed to putting a lot of energy in trying to change policy, which means politics, right. So that’s something else I had to consider that I had to toy around with and make a decision. Including now I was asked to be a director of this or X, Y and Z, because now I’ve made a name for myself in science. Maybe I could or help change. You know, you can’t do it alone. But I’m really still, even at this time in my career I want to make those discoveries about how the world works, how the brain works.

Grant: [00:32:31] What are your thoughts on how people should think about changes that need to happen in the culture of science and so on, right? Are there things out there that you think are especially productive? Things you think are counterproductive? How do you think about it?

Erich: [00:32:45] Yeah. I think what’s counterproductive is for the scientific community as a whole and individual institutions or departments to expect that the folks who are–I don’t want to call them victims, but who are being negatively affected by discrimination and so forth–are the ones who should be given the keys to try to find the solution. You know, I think the solution to what I call society’s racial disease is everybody needs to be get involved, whether or not they are perpetrators or racism or benefit from racial discrimination and so forth, we all need to be part of the solution for it to work.

[00:33:40] The other is that the scientific community needs to do more scientific research on not just social research, right, where you’re looking at behavior only.

But what is it in our human behavior that leads to tribalism in the form of racism? Why does that happen? And is there something genetic about that? Or is there some type of nature versus nurture influence of your social upbringing that leads some people to become white supremacists and others to become activists against those white supremacists? I think there needs to be more hard science that goes into this to find solutions.

Grant: [00:34:07] What do you think are the most important questions to answer along those lines?

Erich: [00:34:12] I think some of the most important ones are where is the overlap or the interaction between tribalism, economics and health? Ok, so I’ll say it again tribalism, economics and health. Because I think those three together are the problems that are contributing to this racial disease where the economics and the health become a problem, right, then the tribalism breaks out. And so where is that tipping point? Then we can find ways to prevent that from happening.

Grant: [00:34:44] What are the ways that you kind of approach this problem that you think differ from maybe how others may?

Erich: [00:34:51] What I think I do that’s differing from many others–I won’t say many others, but enough others–is I think I learned how to have more resilience than I realized. It’s a level of resilience that I don’t think should be necessary, but it was necessary. And that resilience is not only resilience to folks who say or do things that would be purposely discriminatory. Some people really thought I had an unfair advantage with affirmative action programs, for example, or that I am less than or because of the color of my skin, my ethnicity, that I’m not as smart. I’ve met people who think that way in the sciences, right. So you’ve got to have resilience to that. And you’re going to need resilience to implicit bias where someone is saying or doing something, but they don’t realize what they’re saying or doing is discriminatory and they have all good intentions.

[00:35:50] And I say that because I have had enough people in my office, people of color who come to my office crying about something or feeling less than. And I’ll say resilience also to the impostor syndrome. Do I belong here? Do I belong at Rockefeller? I’ve had that question both as a student and interviewing as a faculty member. You’ve got to have resilience to your own imposter syndrome as well.

Grant: [00:36:15] And that’s interesting because at least most recently, right, when you returned to Rockefeller, you were already very, very well established at Duke, right. And even at that time, was that still kind of a something you were dealing with?

Erich: [00:36:27] Yes, it was something that I was dealing with. And it even surprised me when I realized that. And it was a few other famous scientists basically saying that, Erich, you’re lowering yourself too much. And I was surprised because I would rather underestimate what I am, than overestimate. But, some people were saying that I was doing too much of an underestimation and I realized it was because of my minority status that I was doing that to myself.

Grant: [00:36:57] And so how would you encourage earlier career scientists of color?

Erich: [00:37:02] Yeah. I found a way to evaluate myself because like I said, you don’t want to overestimate either. If you get too confident, you might do something stupid in the field and send a grant proposal in that’s really horrible. And then get a reputation that you’re sending in horrible grant proposals for your work. I try to balance my evaluation with my own self-evaluation and what others think. In the beginning phase it’s going to be hard because you’re just starting. So I would say to the younger scientists, and that includes the students, you know, go get some opinions of different faculty members. Don’t depend on one because especially as a person of color, you might find one or two that are going to undervalue you anyway.

Erich: [00:37:46] And if they start saying to you that it’s okay that you’re not going to achieve as much as somebody who’s white, be careful because they might undervalue you. That’s why I say get multiple opinions. Don’t accept everything they say, but listen, understand everything they’re saying and try to improve what you’re doing based upon that knowledge. Later in life as you start to publish papers and so forth, the way I do a self-evaluation is at the end of every year, I see what the citation rate of my papers are, what impact my papers are having on the scientific field. And now I have out of the hundred and sixty or so papers we’ve published over my career even since I was an undergraduate student.

[00:38:33] Something like 20-25 of them, maybe more than that are cited in the top one percent of papers in their field, according to the metrics. So that can’t be because of the color of my skin, you know. So I use that as a metric to answer that question.

Grant: [00:38:53] That’s fascinating. Yeah, we actually, on our blog have compiled, to my knowledge, the only database of H-index distributions at different institutions for biological scientists who involve computational biology in their work. And so you can go and say I’m an assistant professor, associate full professor at this kind of institution, like what the H Index distribution looks like.

Erich: [00:39:21] Wow, that’s good to know. I’ll check it out.

Grant: [00:39:24] Cool. So what do you think has changed about how science is done today from when you were, just entering the field?

Erich: [00:39:34] I was born in 1965, right. And so I entered science when I was eighteen years old, right. So we’re talking early 80s. So when I started out in the early 80s, science was a lone ranger approach. Especially when I got to graduate school, I was taught, you have to figure out everything yourself. And you have to be first or last author on the paper. You know, I mean, this kind of thing is still the same now, but it’s less so than it was before. And I found that that was a Western European model of thinking because I grew up in an African American family, with some Native American culture mixed in there. And of course we were surrounded by European culture around this right. I was thinking of like a Martin Luther King approach, you know, bring everybody together be very collaborative.

[00:40:21] And the advice that I was getting is, I’m doing too much collaboration. That I’m not distinguishing myself enough and so forth. Even from when I was getting tenure and so forth and some other people pulled me aside and whispered, don’t pay attention, you’re doing just fine. So I took this more collaborative approach to science, and now I’m finding that I’m good at it and I’m leading large international consortia for genomics or neuroscience and so forth. And I’m getting credit for it. We produce more papers and we switch around authors on different papers and so forth. And they’re coming out as some of the most highly cited papers in the field.

[00:40:58] And also for me, what really counts is not so much your credit, but the discoveries that are made. And so that was my Malcolm X training, right, which is by any means necessary. So if you need to collaborate, do it. If you need to take the lone ranger Western European approach, do it. What is the necessary approach to make the scientific discoveries? And what I’m finding is that as science diversifies more and as big data in the biological science grows, collaboration is necessary. This lone ranger approach is becoming less and less viable and prevalent.

Grant: [00:41:37] And there seem to be a lot of people who are unhappy with that. You see a lot of complaints about money that goes to the consortia, even though their data are indispensable, right. At the CRO, we rely on GTex, we rely on TCGA, we rely on ENCODE and so on, right. We rely on these big consortia data all the time, every single day.

Erich: [00:42:00] Yeah. I’ve heard people who aren’t into big consortia projects. They just want to have six people in their lab and that’s it. They complain about our consortium projects, about how competitive they are.

Grant: [00:42:14] What do you think is done poorly right now in science? What do you think most needs to be improved?

Erich: [00:42:24] I think two things right. One of the biggest things we poorly do in science is communicate to the public about what we’re doing. I think the public is undereducated in the sciences and sometimes miseducated and purposely so for political reasons. This is something that is changing. When I jump back to the 80s and 90s, we were taught don’t talk to the media that much. Don’t do a podcast, right. Because you’re selling yourself or you could say something wrong. Or somebody could misinterpret it because they’re not a scientist. And therefore, you get a bad reputation in the scientific community. Or you are trying to make a big name for yourself, like Carl Sagan.

[00:43:08] But I think that attitude does us a disservice. I think when it comes to the public good and the scientific community throw the humility out the door, educate the world as to what we’re doing and learn from them as well. So that’s one. And the other is, I think we don’t have a big enough financial investment in the sciences. There is a lot of money going into, you know, I guess the one that get criticize all the time is the military, right. But I think the business world and political world could invest more money into science education, even if those students don’t become scientists. I think it’s going to be better for whatever they go into and for the scientists themselves. Of course, that’s self-serving since I’m a scientist. But I think we can we can do a lot more for the climate turning around what’s going on, climate change for our own health and so forth. And just for basic understanding of the world if we invest it more.

Grant: [00:44:05] Totally agree. And to your first point, the reason I went into the life sciences in the end was actually reading some pop science books that Richard Dawkins had written. And then when I went to grad school and was classmates with the number of people in his department and so on. I learned that within the department, people would complain about him a lot and really it sounded more or less like they were jealous that this guy who gets all this attention. But you know, they’re publishing better and all this stuff. And my thought all along was well, but yeah, OK, maybe his publications aren’t as impactful internally. But big picture, he’s probably having a lot more impact because he’s drawing a lot of people into the field who otherwise may not have gone.

Erich: [00:44:46] Right, exactly. So we need we need people like him.

Grant: [00:44:51] Cool. Do you have any final words for our audience, words of encouragement?

Erich: [00:44:57] You know, pigging backing off this last topic. I feel that what scientists need to learn how to do, including myself, is to translate. Not only to translate from bench to bedside type of translation of discoveries, but translate knowledge from the scientific establishment to the general public. And we don’t have enough good translators. And so it’s good to have you as a translator. But we scientists, we need to learn how to make that vocabulary and grammar and so forth understandable.

Grant: [00:45:34] Totally agree. Well, thank you so much for joining us. It was a lot of fun.

Erich: [00:45:39] You’re welcome.

Transcript of Episode 42: Mauro Calabrese

Grace: [00:00:00] Welcome to The Bioinformatics CRO Podcast. My name is Grace Ratley. I’ll be your host for today’s show. And today I am joined by Mauro Calabrese. Mauro is Associate Professor and Director of Graduate Studies in the Department of Pharmacology at the University of North Carolina, Chapel Hill. Welcome Mauro.

Mauro: [00:00:14] Hi, Grace. Thanks for having me on the podcast.

Grace: [00:00:17] Yeah, we’re happy to have you on. So can you tell us a little bit about your current research on long noncoding RNAs?

Mauro: [00:00:25] Yeah. So broadly speaking, we are trying to understand fundamental mechanisms through which our genome is regulated with the understanding that by defining those mechanisms, we’re going to learn a lot about the basic biology that goes on inside of our bodies. And also we’re going to learn about really important events that give rise to and sustain human disease. So what we study in my lab are these molecules called long noncoding RNAs, unlike a typical messenger RNA that encodes information for protein. These RNAs themselves are sort of the end product. Our genomes, the mammalian genome makes lots and lots of noncoding RNA, like billions of base pairs of it, actually, the majority of which we really don’t know what its function is. It may not have a function or maybe it does.

[00:01:16] And we know from a few really amazing examples, these are genes that people have discovered now upwards of 30 years ago, that at least a subset of long noncoding RNAs play really incredible roles at gene regulation. I can give you a specific example. It’s an essential gene, it’s an RNA, a piece of RNA that’s expressed from the X chromosome in all female mammals. And the function of this RNA is to turn off one X chromosome in every cell, essentially for the life of the organism. So this piece of RNA can transcriptionly silence a hundred and sixty five million base pairs of DNA for 100 years and a billion cell divisions. That’s really incredible. And we don’t fully understand how it works, which is interesting. And then beyond that, there’s this sort of universe of long, noncoding RNAs that get produced by our genome and we have no idea how they work or what they do or whether they have a function at all. And so we use genomics and genetics and cell biology and microscopy and biochemistry and a lot of computational biology to try and understand how long noncoding RNAs regulate gene expression and develop new experimental and computational approaches that will enable others to do the same.

Grace: [00:02:33] I noted that your lab also looks at these long noncoding RNAs in the context of cancer. Can you tell us a little bit about that?

Mauro: [00:02:42] Yeah, we are not so explicitly focused on cancer, but some of the RNAs that we study do play roles in cancers. But I think sort of the biggest area that we hope to impact, like in regard to human disease and cancer, is a really I don’t know if low hanging fruit is the right word, but there’s a lot of really great genomic data in cancers. We know what RNAs are expressed in different types of cancers. And so I think that area is really ripe for discovery. And the biggest roadblock in the field is we really don’t have an understanding of at the basic level, what’s the relationship between the sequence of a noncoding RNA and its function in the cell. For the listeners that know a little bit about proteins, relatively have a much more sophisticated understanding of how protein sequence relates to function to the effect that you could take a protein that’s never been studied before and compare it to all other proteins. And chances are you might find a piece of protein that was similar to a previously studied protein, and that would give you really important clues as to the mechanism of your unstudied protein. So like if a protein has a kinase domain, there’s a good chance it’s a kinase.

[00:03:58] And those types of understood relationships just don’t exist in the long noncoding RNA field. And so the effect of that is that in a disease like cancer, people have looked at this many different ways, thousands of different long noncoding RNAs that are differentially expressed that correlate with different forms of metastasis, that correlate with different cancers that appear to be, if you like, knock down these transcripts, they appear to have therapeutic effects. And I believe clinical trials started last year targeting a few long, noncoding RNAs through Ionis and possibly some other pharmaceutical companies. So these are like there’s RNAs that get expressed in cells. Some of them almost certainly are drug targets. But we have absolutely no idea how to know which ones we should be thinking about targeting and what they might do in the cell. By sort of focusing our efforts on understanding a few noncoding RNAs whose functions are well known, we’re starting to get insight into what are the building blocks that noncoding RNAs use to encode function and ultimately hoping to sort of develop a framework that will allow us to computationally predict or identify regulatory function in essentially any noncoding RNA. And I think cancer is an area that I’m really excited to move into once we have a better handle on the approaches that we’re trying to develop, because I think there’s a lot to discover in that space.

Grace: [00:05:24] Yeah. Cancer, I think is a really great place to start because you have such amazing databases of genetic information and gene expression and all of those sorts of things. And I was going to ask how you ended up in the Department of Pharmacology, because from what it sounds like, a lot of the work you do is in genetics. So can you tell us a little bit about that?

Mauro: [00:05:44] Yeah. Great question. People ask me that all the time. You know, science is so interdisciplinary these days. I think our work certainly fits within pharmacology, and I’ll tell you why in a second. But easily, we could be in a genetics department or a cell biology department or even a biochemistry department. So everything is just cross disciplinary.

But I didn’t necessarily apply specifically to pharmacology departments when I was trying to get a faculty position, but there was a position that was open in this department. And the former chair of the department, who was Gary Johnson at the time, I think sort of recognized the potential that the things that I was just talking about in regards to cancer, like there are these really tantalizing examples of like, OK, we know there’s a few noncoding RNAs that appear to be drivers of metastasis and probably can be targeted with antisense analogues or even small molecules targeting structure of RNAs.

You know, there’s sort of a next wave of therapeutics that are going to involve RNA. And this was true in 2014 when I started the job, which is I think why I got brought into this department. But of course, everyone can appreciate it now, the power of RNA delivery through these nanoparticles that has saved, hundreds of thousands of lives through vaccines. So RNA is not something that historically has been drugged or a drug. But I believe in the future, the next 10 to 20 years, we’re going to see more and more RNA based therapeutics. And so that’s how our research fits into pharmacology.

Grace: [00:07:17] Yeah. Yeah, it’s so fascinating. And I’m excited to see how the technology that is the basis of the mRNA vaccine for COVID-19 is going to influence a lot of science, because, I mean, like you said, historically, it just wasn’t possible because RNA is so unstable.

Mauro: [00:07:33] It’s unstable, it’s big. And I mean, although I guess the technology that Moderna and and Pfizer are taking advantage of, the idea that you could deliver messenger RNA as a therapeutic even in the 80’s. It just has, I think for reasons that you just said, taken some time to take off. But, yeah, it seems like there’s just a lot of new possibilities. And I also am excited to see what happens.

Grace: [00:08:02] Yeah. And there’s also that kind of awareness piece of it. You know, it’s like I know it existed before, but I wasn’t familiar with it. I guess I can’t say that I’m an RNA biologist, though.

Mauro: [00:08:14] Well, yeah. I mean, I’ve been studying RNA for a long time, and I wasn’t familiar with it either. So that doesn’t mean that, maybe I should have been familiar with it, but anyway, like the awareness is a big deal. And now we’re all aware.

Grace: [00:08:28] Yeah, exactly. How was your lab affected by the pandemic when it started?

Mauro: [00:08:33] Yeah, I think it was brutal for us and brutal for many industries, and it continues to be brutal for many industries. But it’s a lot better for us now than it was. I remember it well, I’m sure I’ll always remember it. Like you heard about this virus and it was in China and then you heard like, oh, my gosh, they’re like shutting down essentially all of China with pop up hospitals at Wuhan. And then and it was like getting closer. Then in the beginning of March, it was like, oh, maybe it’s going to come here. A week later, it was like, okay, it’s here.

There were a few days where a UNC hadn’t shut down, but we knew it was coming because everything else had shut down. So my lab shut down like a day earlier and we’re just like, we can’t handle this. It’s definitely happening. We’re just going to close everything. And so, yeah, we just stopped research, just pulled the plug on it. And we were out of the lab for three months. But UNC opened up in June of 2020 with masks at 50 percent capacity, and that enabled us to at least get back into the lab.

But it was just a mentally extremely challenging year. People that I know, they lost loved ones. Many people experienced extreme forms of mental stress, and we weren’t spared from that in the lab. So as a father of two young kids and my wife also has a job, we had zero child care. So how do you do that? Like not very well. You know, it was like really hard on everybody, it was very hard on my family because our kids got pulled out of school and my son is like six months old and my daughter is three and a half and somebody needs to pay attention to them because they’re kids and they need that. So it was a wreck.

[00:10:13] We were exceedingly careful in the lab and we managed to make some progress during the year, but it was really limited relative to what we would have expected. And things began to come online once we all got vaccinated. But it’s still a challenge. I mean, I think we need to be wearing masks at work. I’m in a private office, so I’m not wearing a mask right now. But everyone that’s in the lab wears masks all day. So I think people do more work at home than they would otherwise, which I think is fair. You know, your face gets hot. So it’s a fact.

So we’re still like kind of a little fragmented as a lab because we’re doing more remote work than we were used to. The remoteness of Zoom is enabling on one level, but it’s also stifling. And I think it limits the creativity that we get from being all together for a full eight hours a day. And so we still haven’t gotten that back. And I don’t know whether maybe in a year or two, when we stop having to wear masks at work, we’re making discoveries, we’re making good progress. We’re figuring things out that are interesting and important. But I don’t think we’re operating at the same level that we were before. But we’re close.

Grace: [00:11:31] It’s good to hear that you guys are rebounding a little bit now, although it does make me worried, looking at a lot of the numbers, especially around North Carolina. I have a lot of friends going back to school and everything is increasing again. And it does make you a little nervous.

Mauro: [00:11:47] It does. You know, I think I totally agree. I guess the flip side is that UNC research operations did a pretty good job last year. Of course, the Delta variant is different than it was: it’s far more contagious. But I don’t think there was like a single case where they could say there was workplace transmission at UNC at all last year. Like in all the school of medicine, people are wearing masks, they’re adhering to it. And over the last year, I think objectively it was a very safe place to work.

And so even though the Delta variant is like 10 times as contagious as the original coronavirus, I think it still remains like a pretty safe place to work. For now, fortunately, most of us are being vaccinated. And so I’m not immune to getting COVID, but it’s less likely that it’ll be severe. And so this time, even though cases are similarly high now and probably will be higher than they were in January of last year soon, I don’t think a lot of that is happening at work at UNC or even in classes. I think it’s the personal spaces where people are relaxing and interacting closely where the bulk of the transmission is.

Grace: [00:12:58] Yeah. Yeah, very true. Very true. Those darn college students partying. Yeah, it’s been a hard year, hard year and a half. It’s been interesting to see how the pandemic has sped up science in some ways, but also how it’s slowed science down in other ways. It’ll be interesting to see how it works out going forward.

Mauro: [00:13:22] Yeah, but I think I’ve always enjoyed what I do, and I don’t necessarily do it for the benefit of public health. I just I think I’m interested in it. But I do firmly believe that there’s a strong benefit to humankind, as well as an economic benefit to research. And I think the pandemic on one level is inspiring for what you just said/ Being able to develop these vaccines in a record amount of time, and they’re safe and they’re highly effective. And all the amazing research that’s going on in regard to the coronavirus, I think is really underscored the necessity for science and its power.

Grace: [00:14:00] It’s been great to see all of the collaborative efforts, industry and government and universities working all together. And it’s kind of like a world war. You know, it’s like everyone, all the industries, all the people are coming together to fight, not an enemy, but a virus.

Mauro: [00:14:17] Yeah, I think that’s when we’re at our best, when we come together.

Grace: [00:14:21] Yeah. So tell me a little bit about how you got into RNA biology and how you became interested in these long noncoding RNAs.

Mauro: [00:14:31] Happy to do so. It’s a little bit by chance. You know, you make these decisions over the course of your life and then the next decision follows. So I guess I was interested in gene regulation in college, like I found it to be like just really interesting. There’s like all this DNA and it gets read in different ways in different cells at different times. And that was like really interesting to me. And then I went to graduate school at MIT and when you’re first year of graduate student, you rotate in different labs and pretty much by dumb luck I rotated in Phil Sharp’s lab.

Phil Sharp won the Nobel Prize in 1991 for the discovery of splicing and has just made all these really incredible contributions to our understanding of biology and RNA. And I knew I was interested in mammalian gene regulation. So I rotated in Phil’s lab and it was really great. So I ended up joining that lab. And so that sort of left me cultivate my interest in RNA. And when I was in Phil’s lab, we were studying micro RNAs, which are short, twenty one to twenty three nucleotides long, but they’re like genes. Some of them are extremely conserved, incredibly, even though they’re teeny, tiny and short. And they do amazing things like RNA tends to do.

[00:15:51] So we were studying microRNAs and then I became aware of some of the work that was done by researchers, in at that point, the nascent, long noncoding RNA field. So this RNA, I’m not sure I mentioned it by name, but I referenced it very loosely at the beginning of our conversation called Xist. The function of this RNA is to turn off the whole X chromosome. There were a lot of really interesting breakthrough studies on this Xist long noncoding RNA when I was a graduate student, and those were pretty cool to me. There were also independently some breakthrough studies on a different long noncoding RNA that has a function that’s analogous to Xist. All these RNAs have weird names. They’re just gene names. But this is called Air. Actually, we study it in my lab now. It is this amazingly strange RNA. It’s huge. It’s ninety thousand nucleotides long, which is like crazy. It’s un-spliced, highly unstable. But the function is to silence gene expression over about one third of mouse chromosome 17. And it’s not even conserved outside of rodents. And so it’s got this incredible biological activity that has appeared to have evolved very recently in evolution.

[00:17:04] Anyway, some really amazing work from a researcher in Austria who has unfortunately since passed away, Denise Barlowe. But I remember reading some of her papers, as well as these papers from the Xist field, while I was a graduate student at Phil’s lab they piqued my interest. And ultimately I decided to pursue that area of research for postdoc, I thought it was really interesting. And so I did that. I was a postdoc here at UNC and Terry Magnuson’s lab in the genetics department. I wasn’t necessarily set on starting my own lab, but I was just doing what I thought was the next best thing, which was to pursue a postdoc in an area that I found to be interesting. And I still found it to be interesting at the end of my postdoc. And I had enough success to convince the pharmacology department to hire me. And so that sort of brings us to the present day, I guess is an abridged version.

Grace: [00:17:55] Yeah. In your lab, you have such a variety of approaches and computational biology, microscopy. How did all of that come together?

Mauro: [00:18:05] Yeah. I think you just sort of do what needs to be done and get into it a little bit at a time. I was a graduate student right at the inflection point of the sequencing revolution. When I started graduate school we were sequencing micro RNAs by hand. We would like take these little bacterial colonies and after like three weeks of work, you would get three hundred sequences back. And then all of a sudden there was this instrument that came online from a company that was bought by Roche and then I think has gone under. But all of a sudden, instead of for three weeks of work, instead of getting three hundred sequences, you would get ten thousand sequences, 300 times more. And then like a few months later, instead of ten thousand sequences, you could easily get 10 million sequences. So like in the span of nine months, it was like the genome revolution. And so I learned computational biology for that.

Then I was a postdoc in a genetics lab. And questions arise and you want to figure out what are the most important questions to answer and what are the things I need to understand to answer these questions? And if they’re interesting enough and if they’re within your capacity, then you learn the skills necessary to answer them. Like I’m not a math person. So structural biology is probably something I would never be able to pick up myself. But you just sort of over time, I think, figure stuff out.

Grace: [00:19:25] Yeah, certainly. So we work a lot with biotech companies and generally, people who have such varied experience and who have a lot of random skills tend to do very well in biotech companies. Have you ever considered joining a biotech or starting biotech?

Mauro: [00:19:42] Yeah, I have considered it. The benefit of working in an academic lab as long as you can convince somebody that it’s important, they should fund your research. You can do what you find to be the most interesting. And I think we’ve sort of like fallen into this path. I think, the computational objectives that we have, like we’ve started to make some insight into how we can computationally predict the function of a noncoding RNA. And I don’t think we’ve really finished that work yet.

So I’d like us to get there. And I think when we get there, it’s going to be pretty exciting. But I’m not sure that anything that we have done so far or maybe that we will do is–what’s the word–is like IP-able, you know, like it’s information. We’re trying to figure out how to figure things out. And when we figure that out, we’re going to publish it and make it public.

So I think some people’s research, especially in pharmacology is good for biotech. Like we have this protein that we study with a mutation and we’re developing a small molecule to fit in this pocket. And you can then spin off that molecule in IP. We haven’t quite done that work, but I think about it, many of my friends actually from graduate school went into industry, and now they’re like doing all these amazing things at high levels in these biotech companies. And I’m like, wow, that would have been cool. But this was my path. And it’s been great so far. But I’m not averse to doing work in the biotech sphere as well.

Grace: [00:21:04] Yeah. Your path hasn’t ended yet, so you’ve still got time. And if IP-able isn’t a word that’s used in biotech, I think they should add it to the biotech dictionary. What skills do you think are most important for scientists today?

Mauro: [00:21:23] You know, I think a lot of them are skills that are transferable to everything: you know, focus. This is maybe a weird thing to say, but we live in a very distracted time, especially during the pandemic. I mean, the news cycle is like from one crisis to next. And there’s all these things that take our focus away, like text messages and email and different forms of social media. And I think to really do excellent work, somehow you need to put your blinders on and think deeply about stuff, especially if you’re trying to make discoveries. To figure out things that we don’t know yet really takes a lot of ruminating and deep thoughts. So an ability to focus, a strong interest in the work that you do, because that’s going to give you those insights. You’re going to be thinking about your work and you’re going to have the insights, but if you don’t find it so interesting, then you’re not really going to be thinking about it that much. And that’s fine. But it’s not great for research.

[00:22:29] And then I think an ability to communicate. So it’s not enough. It’s never enough, I think, to work in a silos. Our best advances as humankind have come from collaborative efforts. And I think it’s probably always been true, but it’s definitely true today. Especially the more complex a project is, the more likely it is that you’ll need to rely on multiple people to achieve an objective. And at The Bioinformatics CRO–I’m sure you understand this–you have computational biologists and you have biologists who don’t understand computational and communication is really key. Like if you have people that communicate well, then the speed of the project is like orders of magnitude, greater than two people that are not communicating well. So I find that to be extremely true with our work as well. Those are the big ones, I think focus, interest, and an ability to communicate. If you can do those three things, you can pretty much do anything.

Grace: [00:23:30] So the things that you’ve mentioned are traditionally very difficult to learn. So the communication, like team management, that’s very difficult and focus I mean, it’s impossible. I mean, really. So what are some ways that people can cultivate these skills?

Mauro: [00:23:45] Yeah, and I think that’s a great question. And I do think we all need to cultivate them. They’re not necessarily taught in class and they’re not really taught in society either. The way that the news communicates to us is not really the way that we should communicate to each other. The way that like Facebook inundates us with distractions is not healthy. So you really have to like steel your mind against the forces of the world that we live in. To varying success, I have been able to do this. I mean, I think there’s times, especially last year during the election where I was unable to focus. Actually, much of last year, I was unable to focus for a lot of reasons, and I wasn’t alone.

But I think dedicating a time to shut off external communication via text message, take your Apple watch off, don’t listen to music, turn off Slack. Setting aside time to make sure that you do that, I think that’s what I do. Whether or not that’s a transferable skill, I don’t know. But I find that to be very helpful. The times when I’m able to not check my email every five seconds and not look at the news every five seconds. Like those are the days that I feel the best about the work that I’ve done. And those are the days that I’ve accomplished the most.

[00:24:59] I think, similarly in terms of communication, that’s something that I’ve learned. Writing in particular in my job is essential, and it’s not anything that I was formally trained in. I took like an English class in high school, and I learned that a paragraph has a topic sentence and three sentences after that support the topic, which is true sometimes, but that’s not actually the best way to write most of the time. But there is a structure and understanding what people expect and how to include it in a document and how to write words that are going to give you the highest probability of success communicating to your audience. That’s something I’ve read about, I mean, by necessity. If I can’t write, I don’t have money to fund my research. So like anything, essentially, you have to sort of set a goal and work towards it in practice and seek out resources that can help you. And in terms of communication, if you’re a natural communicator, then you have a leg up. But if it’s something that you want to improve, there’s leadership courses, there’s books that people would recommend, I think, in practice and engage and self evaluate. And over time, you will improve.

Grace: [00:26:11] So as we wrap up this episode, I always like to ask our guests what advice they have for early career scientists or graduate student postdocs who are looking to go into a similar path that you’re following. What are some of the lessons that you’ve learned and words of wisdom that you might share with younger scientists?

Mauro: [00:26:33] Yeah. You know, the thing I tell many of the graduate students I interact with and I would share here, is to do your best to figure out what it is that you are interested in. And if you’re going to join a lab or do a postdoc or pick a research project or join a company that you find interesting and that you believe in the mission. Like this work is interesting to me and I believe that it’s worthwhile. And I feel like that’s like the number one thing.

I couldn’t tell you why I’m interested in the work that I do. I just find it interesting and I don’t think I need to justify it. It’s interesting to me, and I can tell you why it’s important. And so I think that’s true. I suspect for other people, like you don’t have to justify it to anybody. It’s meaningful to you and just tap into that feeling, like don’t overthink it. But if you find yourself feeling like, it’s a slog–it’s bound to happen–just sort of reflect on that. And when it comes time to make another career decision, think about the things that you’ve done and what you enjoyed and what you didn’t enjoy. I think follow your interests as best as you can and that’s going to get you really far because it’s the more interested you are, the more excited you are, the better work you’re going to do, the harder you’re going to work and the greater successes that you’ll have.

Grace: [00:27:53] Excellent advice. So thank you so much Mauro for coming on the podcast and sharing your wisdom and your experiences. I had a really great time talking with you today.

Mauro: [00:28:02] Yeah, Grace. It was really fun.

Transcript of Episode 41: Damian Kao

Grant: [00:00:00] Welcome to The Bioinformatics CRO Podcast. I’m Grant Belgard and joining me today is Damian Kao. Damian is the Chief Operating Officer at Basepaws. Welcome.

Damian: [00:00:10] Thanks, Grant. Happy to be here.

Grant: [00:00:11] Happy to have you. So, can you tell us about what Basepaws is? What do you do there?

Damian: [00:00:15] Yeah. So, Basepaws started out as a kind of a 23andMe for cats type of business. I think a good analogy would be Embark for dogs. However, I think in recent years we’ve been kind of pivoting into more of a pet health care angle. So, the goal of Basepaws is really to utilize as much genomics data as we can for our pets to profile their health and try to predict health outcomes, basically. So, preventive medicine in some sense. Right? So, being able to predict whether your cat or dog will have a disease before it happens, so we can save you that huge veterinary bill that you might have down the line.

Grant: [00:00:56] And you mentioned dog. Are you guys moving into dogs?

Damian: [00:00:59] Dog is a very competitive market, as you may know. Embark actually recently just got a lot of funding from SoftBank. I think they’re valued at 700 million dollars now. So, that’s obviously a big competitor. We are thinking about other animals. It’s a bit intimidating to go into the dog space so we might try to go in there with some other products. So, for example, we just came out with a dental/oral microbiome product, actually. Where we give you a risk score on whether your cat will develop periodontal disease or some other health issues. So, we might try to enter the dog market with that type of product.

Grant: [00:01:39] How actionable are those reports? So, if you’re a cat owner and you find your cat is in the 90th percentile of risk for periodontal disease, do you brush their teeth? The way you’re supposed to, right. I don’t think most cat owners do routinely brush their cat’s teeth.

Damian: [00:01:55] Most pet owners in general don’t brush their cat’s teeth, right? Or their dog’s teeth. It’s, as you may imagine, very difficult to brush them.

Grant: Getting your hands anywhere near their mouth is this just asking for trouble I think.

Damian: [00:02:16] Yeah, I think the goal is to at least let the cat owners around the world know that this is an issue. Periodontal disease in cats is a huge issue and huge vet bill. So, there are a series of products recommended by veterinary council that is not always active brushing. Maybe you can give them chews or water additives or certain types of food that might prevent those types of problems. And that’s what we are advising people to do right now.

Grant: [00:02:37] And what kind of dynamic range do you have? You know, so a cat at either extreme, what kind of difference in rescue you’re looking at?

Damian: [00:02:45] So, let me talk a little bit about the bioinformatics of that, I guess. Our product is really just a swab that we provide to our customers. They swab their cat’s mouth, they send it back to us. And from the very beginning, we’ve noticed that after sequencing the DNA up to 10 to 15 percent of the DNA is not cats. So, what could that be? It’s just whatever is in the cat’s mouth. That could actually be residual food. That could be the oral microbiome flora. That has always been really exciting and interesting for us. And we didn’t really act upon it until maybe a couple of months or a half a year ago. But what we did find is we have a cohort of at least 30000 cats. So, 30000 oral microbiome samples. And then we have good phenotype data that tells us whether the cat is on a certain type of diet, whether they’re indoor or outdoor cats or whether they have any systemic diseases. So, this became a really interesting thing for us, because now we can try to look at a microbiome profile and correlate it with all this phenotypic data that we’ve gathered. Dental disease is obviously the most direct thing that we can look at.

[00:03:53] So, in our cohort of cats, we have hundreds of cats with periodontal disease and cats with other oral issues. And we’ve found that certain populations of microbes in combination seems to be correlated to these disease states. We don’t do 16S-rRNA seq. We are doing the WGS. So, we are looking at everything that’s in the mouth. So, we do see fungus. We do see bacteria. We do see some archaea. And we actually see a lot of residual food things, too. We pick up on plants, maybe a spider that the cat ate while it was outside. So, we do pick up on those type of things, too.

Grant: [00:04:30] Do you have a longitudinal component to your data? So you can certainly imagine if a cat has active periodontal disease, that their oral microbiome probably looks pretty different at that point. It would be super interesting if you had good predictive power years in advance to say like this cat is at high risk.

Damian: [00:04:48] Yeah. So, there are a series of studies that we’re actually conducting right now with various clinics all over the country that are gathering these samples for us. So, we’re working with dental specialty clinics that are gathering samples before and after examination, for example, and maybe also follow up weeks or months after examination. It’s through these samples that we want to start looking at what the predictive power really could be.

But the longitudinal question is really interesting for us, too, because we are seeing signal not just for oral or dental issues. We are seeing interesting signals for some systemic diseases, too. And this is reported in literature somewhat. For example, Ckd, chronic kidney disease in cats. There does seem to be some link between that and periodontal disease. And in our data set, we are seeing a signal coming out for chronic kidney disease via our microbiome dataset. We are also seeing some signal for other autoimmune diseases or even allergies, which is kind of interesting. It’s hard to kind of tease all of these things apart, though, right? So, getting good high resolution phenotypic data, I think is really our next big thing.

Grant: [00:06:01] How careful do you have to be about how you ask those questions, right? Because you can imagine if you’re asking about behavioral traits, the same pet owner may describe their cat’s behavior in very different ways.

Damian: [00:06:13] There is definitely an art to asking questions from our customers, and that’s something that we’ve had a lot of trial and error on. I think the best thing to do is to ask them the same question in multiple ways, many times, spread across multiple time points. And that’s really how you can increase the confidence of those answers. That’s something we’re building into our account system right now. The whole idea is we would have a question bank in our backend where we might ask the same question 10 different times, and we would present those questions to the customer at various times. And hopefully they will be consistent in their responses.

Grant: [00:06:51] Have you ever looked at what’s predictive of inconsistency, right? Do you have certain respondents who are just consistently inconsistent?

Damian: [00:07:00] No, we have not looked at that. But that is actually really interesting. Yeah, we should definitely do that.

Grant: [00:07:07] It would be super crazy if you found something different about the cats, right? Maybe certain cat breeds are associated, you know. But this brings up an interesting point, right? Most of the people we have on the podcast work for companies that are essentially B2B or they’re just drug development companies. What are your experiences with running a B2C Science company? Because generally, in tech, people always talk about B2C’s having a lot more challenges.

Damian: [00:07:35] I have never worked at a B2B, so I can’t really make that comparison. However, I could say that B2C has definitely been a huge learning experience for me. So, I would say that you have to balance satisfying your customers and maintaining the scientific integrity of your work. And that’s always difficult because they’re not always in sync.

Grant: [00:07:58] People always want more information. Right? Like what kind of wine do I like based on my DNA kind of thing.

Damian: [00:08:04] Yeah. Because it’s very easy to kind of stretch the science to accommodate what people want to see. And there’s a limit to that, right? It’s always a push and pull, to be honest. So, at Basepaws I’m the COO. I run kind of the science and the tech and the lab operations. And then we have the other half of the company, which is run by Anna, which deals with customers and acquisition and marketing and all that. And I find that it’s a great push and pull between me and her, because I tend to always think I need to be crazy rigorous. We can’t show anything. But she’s always like, if it’s interesting, as long as we explain it well, we should be able to show to our customers. Because we have to trust that they’re smart enough to understand not to take this at face value. So, I think there’s always this push and pull between kind of the marketing side of things, that B2C side of things, and the science side of things. And that’s been a really good experience for me personally, I think.

Grant: [00:08:59] What have been the biggest surprises for you in your Basepaws journey?

Damian: [00:09:04] I think part of the reason why I decided to not stay in academia and go industry was… There are multiple reasons why. I’m sure most graduating PhDs or postdocs will understand the reasons why. But I guess the biggest epiphany I found was that I enjoy building things rather than answering questions. And I think that’s kind of the biggest epiphany I had doing Basepaws. Building up a Lab, setting up these processes, seeing things happen and producing a product is extremely enjoyable.

Grant: [00:09:36] What elements of that do you think you are prepared for through your education and training? And what did you really have to pick up on the job?

Damian: [00:09:44] The advice I can give to any students going into their PhD is learn transferable skills. You’re not there to learn a very specific lab technique that only five labs in the world can do. You’re there to learn how to think. You’re there to learn how to pick up a new skill when it’s presented to you. So, learn those type of skills, don’t memorize concepts, don’t learn some niche technique. I think bioinformatics is very much a field where you don’t learn specific techniques because there aren’t really any standardized techniques for bioinformatics. It’s kind of a Wild West still in some sense. I feel like you really have to understand algorithms very well, data structures, etc. I think all of these things that you have to learn through your bioinformatics, PhD helps you in industry. It definitely helped me, sitting in my lab knowing how to analyze the data that comes out of the pipelines I set up. What I had to learn on a job is really more management skills I would say. In the lab I manage a couple of technicians. I manage the tech team. So, how you get all these different people to kind of share your vision and execute on that vision, that’s very difficult. It’s not something you learned during your PhD. And I had to learn that on the job.

Grant: [00:10:59] Let’s talk about you. So, when you were a kid, did you wanted to be a scientist?

Damian: [00:11:05] I was a very unremarkable student. As a kid, I would say in university, I changed my majors a lot. I didn’t really know that I wanted to be a scientist. I actually liked graphic design. I was a graphic designer in high school.

Grant: [00:11:17] Have you ever used that skill? Has that been one of those transferable skills that came in handy?

Damian: [00:11:21] Actually, the current Basepaws logo, I designed it all and I coded it all. So, there were some transferable skills there. I did film studies for a little bit in college, and I decided to go in genetics because I also studied a lot programming in high school. So, I made my own websites that type of thing. Back then, in the late 90s, that’s what a lot of computer geeks did. And of course, I did that. So, those programming skills led to my interest in genetics, because there are those obvious parallels between programming and genetics. After learning more about genetics, you realize those parallels don’t really apply that much. But I think that’s what kind of made me want to become a scientist: through computer science.

Grant: [00:12:04] You ended up landing on genetics at UC Davis. What did you do after that?

Damian: [00:12:09] I actually did not think about going into bioinformatics. I wanted to do lab work. So, I was a lab technician for a couple of years, working on drosophila. I did a lot of molecular work, did a lot of injecting stuff into Drosophila eggs to make transgenic lines and all these things. After a while honestly, I was a bit lost for a little bit, just didn’t know what to do. At some point, I decided that I needed a change of scenery. So, I said, I should do a PhD. Let’s go to another country and do it. So, I went to U.K. and did my PhD there.

Grant: [00:12:43] What attracted you to the UK?

Damian: [00:12:44] It was a different country. That was a main reason. You know, I felt like I’d been in California for so long. I feel like when you’re in one place for too long, you lose opportunity to kind of reinvent yourself because you’re surrounded by all the things that you know. Going to the UK allowed me to kind of reinvent myself, I guess, to maybe see myself in different lights. And it was kind of there that I developed that state of mind where I wanted to do a PhD. I want to do all of these things. I was able to, I guess, be more aggressive about my goals in some sense.

Grant: [00:13:16] And what are your thoughts on the British PhD training system as compared to the American system?

Damian: [00:13:23] I mean, there’s pros and cons to both. I think the biggest pro for me on a very practical level is that you’re done in four years, five years, max. After that, it’s really looked down upon if you’re not done. The universities I think lose funding if they have PhD students for longer than a certain amount of time. So, they really are incentivized to get them finished. So, that’s practically that’s one of the biggest con. And then personally, the UK is extremely strong in bioinformatics, as you probably know. So, my supervisor and a couple other PI’s around the UK would yearly set up a genomics conference that I would be a part of where I get to meet all the other great bioinformaticians there. And that was a real, really good opportunity for me to connect with others and learn as much as I can about the entire field.

Grant: [00:14:10] And you must have liked it a lot, because after you finished your PhD, you stayed.

Damian: [00:14:14] Yeah, I stayed for a couple of years. Yeah. After the PhD, I thought about staying academia. I worked on some genome assemblies. I worked on some transcriptomic stuff. So, doing a postdoc in Oxford was very eye-opening to me because there are a lot of really, really smart people. And you just learn so many things, new things and interesting things every day. I was in a zoology department where you got to look at other people’s research and a variety of animals and biology that’s out there. I think that’s one of the biggest reasons why I stay around for a couple of years.

Grant: [00:14:49] Why did you leave?

Damian: [00:14:51] I left because I decided I didn’t want to stay in academia and if I didn’t want to stay in academia, I might as well go home. And I feel like the startup culture in the US, especially California, is just more vibrant.

Grant: [00:15:06] And when you returned to the U.S., you started HHMI, right?

Damian: [00:15:10] Yeah. So, through some work that I did in Oxford, I became a consultant at HHMI where I worked on some single cell transcriptomic projects and some genome assembly projects, and that was a really cool experience for me, because Janelia is just a great place to work. They basically built this entire compound where you can live and everyone just loves science and does science. It’s great.

Grant: [00:15:37] Bit of a scientific monastery, right? Isolated and such.

Damian: [00:15:40] Yeah. Monastery. That’s a really good word for it. Yeah.

Grant: [00:15:44] What was your thinking in leaving there? Was it basically you really wanted to do the startup thing? And what’s the story there? How did you and Anna meet?

Damian: [00:15:52] Yeah, I really wanted to do start up things. So, I was in California remote working for HHMI. I just felt like doing the same type of analysis on the same data sets is just going to be boring for me.  And I really wanted to get into the startup world and see how that works. My mom was entrepreneur in Taiwan and she’s a successful businesswoman there. And I’ve always wanted to see what that was like. I think at the end of day, I wanted to work for myself. Didn’t want to work for someone else. So, my wife did PhD with me at the same place. We came to California together. We both are kind of into this whole startup scene. So, we put out some feelers.

Grant: [00:16:32] Is she also American or?

Damian: [00:16:35] She’s actually a Bulgarian. She did her PhD in the same lab, got together there, got married in the UK. So, we put out some feelers in the startup world. I had some NGS experience. Anna was in need of someone with that experience. So we met at a coffee shop one day, talked about our respective skills and our interests, and I thought that we were a perfect fit to do this company together. So, I joined and I set up the lab, did all the pipelines, and we went from there.

Grant: [00:17:10] What were your biggest challenges when you got started with Basepaws?

Damian: [00:17:13] I think something that carries over from academia and into the first few years of industry is imposter syndrome, I think a lot of people have that. The first year at Basepaws you know, I’m the scientist. I know the science. However, the business side was not something I’d experienced. And so whenever I had to make a business decision, I would always second guess myself. And I think it stems from that imposter syndrome. But I think at the end of the day, what I learned is that business decisions are like any scientific decision. You get a lot of data, you analyze it and you make a decision. There’s nothing special about it. So, getting over that imposter syndrome and having more confidence in the decisions that you’re making. Yeah, I think that’s what I learned.

Grant: [00:18:01] Do you think you’ll ever leave the startup world?

Damian: [00:18:04] Well, I just had a kid who’s one year old, so it’s a Covid baby. If I ever leave the startup world, it will be because it’s becoming too crazy and I can’t spend enough time with my kid. That’s probably going to be the reason.

Grant: [00:18:17] It is tough. What advice would you have for people considering doing the startup thing? Like we have a number of clients who started their company after, spending a long career in big pharma, where everything was taken care of for them. And they would focus on their one area and they were the expert on that. But, they could access any kind of expertise they wanted just by going down the hall. And obviously that’s not the case at the start up. Right? So, what advice would you have for people like that considering making the jump?

Damian: [00:18:51] I think hugely depends on what your business is. I can tell you from a kind of B2C point of view that scaling up is very, very hard, especially in a biological context. It’s very easy to get an assay or a product or a test to work a couple of times, but to get it working consistently for thousands or tens of thousands of times, that’s extremely hard. So depending on what industry you’re in, if you’re in an industry where you have to do that thousands and tens of thousands of times, you have to think about that. And you have to think about the long term cost of what you’re doing, because it’s also very easy to over-optimize and over-engineer in the beginning, buy all this fancy equipment that you just never use because there’s simpler solutions out there. So, I would say worry about scaling up if that’s the industry you’re in.

Grant: [00:19:44] It brings in a very interesting point. The funding climate right now for human therapeutics is quite hot. How well does that translate to pet health?

Damian: [00:19:56] Pet health is actually one of the industries that grew a lot last year during Covid and is steadily growing, and because of that, there are actually plenty of investors interested. I think the problem that stops an investor from actually putting in the money at the end, though, is that there haven’t really been any big exits that they can see in this sector. So, I feel like maybe that’s what’s holding it back a little bit. So, I think there’s a lot of money pouring in. And because most of these investors are looking for a relatively quick turnarounds, they’re a little bit more hesitant to put their money in.

Grant: [00:20:30] Can you tell us about how you use bioinformatics at Basepaws?

Damian: [00:20:35] Yeah, so just a brief overview of what we do. I think we’re actually one of the few companies that use NGS for this type of a product. I mean, as many of you may know, 23andme and Embark and these other companies, they all use microarrays, which relies on an already good existing resource for that organism, like humans and dogs. When we first started Basepaws, a really good feline genome was produced actually a couple of years back. So, we were able to take advantage of that. However, in terms of whole genome data sets for cats, I think when we first started, there were less than 100 in NCBI. So, we had to sequence a lot of those things ourselves, build up our own reference panel, our own imputation panel.

Grant: [00:21:20] So, that you could create a crazy cat assembly. This one, if you have 30000 cats at this point.

Damian: [00:21:25] Yeah. I mean, that was a project that we were thinking of doing. However, there was a really interesting paper as it came out earlier this year where they were able to a haploid assembly of feline genome by sequencing a wildcat and domestic cat. And then the offspring. So, from those reads from the offspring, they were able to say, all of these reads are the domestic cat, all these reads are the wildcat. And then they did a pseudo haploid assembly using that method is really cool. So, that’s a really, really good genome. And I don’t know if we can beat that, to be honest. One interesting thing we are thinking of doing is to try to produce a haploid stem cell line, because that is something as possible to get an oocyte out of a cat and use strontium chloride to kind of activate it. And it can sometimes become a haploid stem cell line. And when you have that then you can sequence the genome and it’s a haploid genome.

[00:22:17] So, you don’t have to worry about heterozygosity or any of that. So, I Basepaws we do low-pass sequencing. We sequence at around 0.5-1x. And then using our imputation panel, we impute a lot of other markers. Usually, we end up with a couple of million markers at the end. And then using these markers, we use a machine learning algorithm. We just use the random forest-like algorithm really to assign haploid type segments to a known breed. And then we calculate a similarity of your cat to this breed. So, that’s what we do for our breed analysis portion of things. And for the health markers side of things, we have a multiplex amplicon panel that we’ve developed where we interrogate I think right now, 40 or so loci using this multiplex amplicon approach. And then we give you the status of whether there are heterozygotes, whether you have copy. How many copies you have, that type of thing. And we are expanding that panel to about 120 markers by the end of the year.

Grant: [00:23:19] What conditions have you found good predictive power for?

Damian: [00:23:23] So, this is the stuff I get excited about, right? Because I feel like when people talk about bioinformatics, they have this artificial separation thing, genotype and phenotype. I think the correct view of looking at it is it’s just all data. It’s all just some kind of dimension of the sample that you collect. And I think when you throw all that together into multi-omic analysis, that’s where the power comes in. So, that’s what we’re working on right now. So, like, you know, like I was saying that CKD, the chronic kidney disease signal that we’re seeing from the oral microbiome, we’re seeing a big signal from that. However, it overlaps with the periodontal disease signal. So, it’s hard to tease apart. Does this cat really have chronic kidney disease or does it just have periodontal disease? However, if we apply a layer of genomic data or some other phenotype that we get, we find that we can pry these apart a little bit more. So, we’re still trying to find the set of features that can best tease those things apart. But I think we’re getting close to some interesting set of features.

Grant: [00:24:20] And how translatable do you think your findings in cats will be to, for example, human health?

Damian: [00:24:26] So, there was a great review paper by Leslie Lyons, who is kind of the main person in the feline genetics field. She wrote a review talking about how if you compare the feline genomes to the human genome is actually one of the closest mammals that exists. I think it’s the most syntonic aligned genome compared to every other mammal out there. You know, if you look at something like genes involved in eye development, I think all of those genes are syntonic with the human block of genes. So, I think there’s a lot of translation potential by studying felines. And I think a lot of the known health markers in felines have almost a direct homologous variant in the human genome, too.

Grant: [00:25:08] Is that something that Basepaws is planning at, looking at in a systematic way?

Damian: [00:25:12] It’s one of my pipe dreams, to be honest. I mean, there’s so many things we can do, but I think maybe like five years down the line, whatever it is, let’s say we collect a ton of cat data. We collect a ton of dog data. You know, can we have a pan-mammalian database where we just like all the variants and use that to narrow down disease markers? So, in humans, you find 80 potential markers for diabetes. You get to narrow that down to 10 because you find homologous variants in these other animals. I think that’s a great usage of this data.

Grant: [00:25:41] How do you think about R&D and kind of building capacity versus having a sustainable revenue driven company? Because generally in the biotech space, most companies are pre revenue for a very long time. And obviously Basepaws already is very actively engaging with customers and has been for a long time. But at the same time, you’re doing a lot of internal R&D work. So, kind of what’s your framework for that?

Damian: [00:26:13] In terms of R&D I always separate in two buckets. One is maintaining or optimizing what we have currently. That means lowering costs for library preps, how we can normalize things better. And the second bucket is what new products we can get from that. In terms of new products, I think for the last one or two years, we’ve mainly been focused on the bioinformatics side of things because it’s cheaper. That’s really the reason we have a lot of data. Can we generate new products from that data? Which we have with the oral microbiome product, for example. But I think now we’re actually close to the end of our series, a funding. I want to start focusing more on kind of lab assays or products or tests that we can do. Something I’m kind of interested in doing is one of those epigenetic clock aging test type of things.

Grant: [00:27:02] Oh, that would be cool.

Damian: [00:27:03] Yeah. You know, one thing I’m kind of interested in figuring out, and there are a couple of papers on this already, is the DNA that you get from on saliva, does that correlate with the blood DNA that’s traditionally done in the epigenetic clock studies. There are a couple of papers looking at that and saying that it does correlate. So, maybe you can do all these epigenetic aging tests through the saliva DNA. That would be cool.

Speaker3: [00:27:26] Very cool. Very cool. How would you think about the kind of commercial path for in, you know, a cat saliva based epigenetic test of aging?

Damian: [00:27:36] I think biological aging is something that people are just interested in. And being able to gather that data and compare it to its real age can give you a lot of insight into longevity and a host of other interesting biological concepts. Longevity is something that me and my wife are personally interested in. So, I think a lot of people might be interested, too. And we’re always looking for products that are not the standard 23andMe ancestry or health marker type of tests. And this is just another one of them, because I feel like if we want to enter, especially a dog space or other animals, you need to have something different.

Grant: [00:28:12] How do you think about engaging with your community?

Damian: [00:28:14] So, the cat community is very different from, as you might imagine, other dog or human communities. People are a lot more obsessive about their cats, I would say, in a good way. I don’t want to suggest that that’s a bad thing. And I think they are a lot more curious about their pets than they are about themselves, actually. That’s one trend I’ve seen. I wonder about this in the human space, too. I would much rather get a DNA test for my kid than for myself because I think most people are like, oh, I know myself, I don’t need to know more. So, I think in the pet space, that kind of applies too. I would much rather find out more about my cat, my dog, who can’t really tell me what’s wrong, than about myself. I think maybe that’s one advantage we have over the human space in some sense.

Grant: [00:29:01] Great. Is there any advice you’d have for scientists looking at transitioning into the biotech startup world?

Damian: [00:29:09] I think as an academic scientist, I don’t want to paint the situation with a broad brush here. But I think the academic mindset, sometimes it’s like I have a choice. I can do really good science or I can have enjoyable personal life. You know, I think it’s a false choice. Personally, I think you can have both. When a scientist gets into an industry, they maintain that mindset a little bit. And I think industry sometimes will try to take advantage of that. So, I think any academic scientists going into industry should change their mindset. They should see their value, get over that imposter syndrome and know that you’re probably one of the few people who can answer or solve these types of problems, have that confidence. I think in academia, when you’re surrounded by a bunch of really intelligent people, it’s kind of hard to have that kind of confidence. I guess, don’t carry over your academic baggage into industry would be my best advice.

Grant: [00:30:05] Right. It’s like a lot of really smart people who enjoy shooting each other down, right?

Damian: [00:30:09] Exactly. Yeah. Yeah.

Grant: [00:30:12] Do you have any parting words for our listeners?

Damian: [00:30:15] I mean, since this is The Bioinformatics CRO Podcast, I would say that I’m excited about the future of this field. There’s a lot of interesting things happening. And I would encourage more people to join this industry because there is a lack of bioinformaticians. We’re hiring, by the way. So, apply for a job with us, if you’re interested.

Grant: [00:30:36] Thank you so much. It was great having you on.

Damian: [00:30:38] Yeah, no problem, Grant. Thank you.

Transcript of Episode 40: Nicholas Heaton

Grace: [00:00:00] Welcome to The Bioinformatics CRO podcast. My name is Grace Ratley. And today I’m joined by Nicholas Heaton, who is Assistant Professor of Molecular Genetics and Microbiology at Duke University. Welcome, Nick.

Nick: [00:00:11] Thanks, Grace. Great to be talking today.

Grace: [00:00:14] Yeah. So let’s start off with talking a little bit about what you study at Duke, and that is the influenza virus and other things occasionally. Can you tell us a little bit about that?

Nick: [00:00:26] Yeah, absolutely. So our lab, you know at the highest level is kind of broadly interested in understanding how these viruses make you sick. We basically modify the virus itself so that that virus will act as a tool that we can then use to ask whatever kind of scientific questions that we’re interested in.

Grace: [00:00:45] What kinds of things in particular do you study? Are you looking more at the genetics aspects of influenza viruses or at the structure or how they infect cells?

Nick: [00:00:55] Yeah. So a lot of the things that we’ve been working on, we kind of think about them as falling into the margins or like the gray space of what happens during a viral infection. So for the most part, when a cell is infected by a virus, scientists think of like a program being initiated. And the same thing essentially happens over and over and over. And what we found and others have appreciated as well, is that it’s more complex than that. The different cell types can respond different ways, and that happens at various frequencies. So that’s what we’ve been studying. The more rare outcomes of infections and trying to understand the nuances of how the virus can interact with its host can really then go on to dictate high level phenomena like disease severity or transmission or something like this.

Grace: [00:01:44] And could you tell our audience a little bit about influenza viruses? I mean, most people have heard of flu season and things like that, but what are some basic virology things that people may not have heard about?

Nick: [00:01:56] Yeah. So these viruses, they fall into a family they are called Orthomyxoviridae. And it’s a huge family of viruses. A lot of them are insect viruses, actually. But the ones that fall into the influenza virus family are influenza A, B, C and D. And when we talk about flu, like clinically people getting the flu or mostly talking about influenza A and influenza B. In a given flu year, it’ll be about two thirds influenza A and a third influenza B. Sometimes those ratios can flip. Influenza C’s can infect children sometimes, but they have animal reservoirs. And influenza D’s are rarely, if ever, detected in people. But these viruses, they have genetic material that’s encoded in RNA.

[00:02:31] So influenza viruses are called RNA viruses. And they’re envelope, which just means they take essentially part of the cell membrane with it when they leave. And that’s how they essentially incorporate it into an actual physical particle that can be sneezed on somebody or breathed in. And these are respiratory viruses like coronaviruses or other viruses.

Grace: [00:03:01] Awesome. So where do they originate? Influenza is a zoonotic virus, if I’m not mistaken. So what’s their normal reservoir?

Nick: [00:03:10] Yeah. So the answer is that it’s complicated. The family of influenza virus especially, so influenza A’s, which, you know are kind of the predominant flu strain. It’s really a bird virus. So there’s lots and lots of different subtypes of influenza that are all found in birds. Migratory waterfowl is where you find these things. And that’s probably almost certainly in the initial kind of introduction of those viruses into people or into mammals. But now we just have human strains of flu that just spread from person to person. It’s not like it goes from person to bird, back to person to bird. Now, we have human strains, and our human strains of flu are to a large extent still related to those bird viruses. But yeah, now they just circulate in people. And the subtypes I mean, the letters and numbers that you hear are like H1N1, that refers to the subtype of the virus. And so H1N1 and H3N2’s are what actively circulate in people right now.

Grace: [00:04:08] And for our listeners, the H’s and the N’s refer to the proteins that are found on the envelope of the virus, is that correct?

Nick: [00:04:15] Yes. On the very outside of the virus there’s two kind of dominant proteins:  H is the hemagglutinin protein and N the neuraminidase protein. And that’s where the H and the N come from. And basically there’s dozens of different types of hemagglutinin and different types of neuraminidase’s. So H1N1 that I referred to, that’s a subtype one hemagglutinin and subtype one neuraminidase. And again, these are just the proteins that are on the outside of virus that actually help the virus get into the cell. And because that’s what the immune response is mostly directed against, people have characterized them in these groups based on reactivity of antibodies, is essentially how the subtypes are defined.

Grace: [00:04:56] Yeah. So talking a little bit about how we prevent influenza infection every year, you know, we have our seasonal flu vaccine. But there’s a lot of, I don’t want to say controversy, but I feel like a lot of people are really reluctant to get those vaccines. And I feel like a lot of that has to do with people questioning the effectiveness of the vaccine. So can you tell us a little bit about what it means to produce a vaccine for influenza?

Nick: [00:05:23] Yeah. So we’ve been making vaccines against influenza viruses for a long time. So there’s different kinds of vaccines now. The majority of them are what we call split vaccines, which are viruses that are grown and then purified and then treated with a detergent so that they fall apart. So you just have all the pieces of the virus, but there’s no infectious particles there, and that’s what’s injected into your shoulder. Like I said, the vast majority of people get vaccines, get those. There are also some purified protein vaccines where just the virus is never involved. You just express different proteins from the virus that can be injected. There are some live attenuated vaccines, those are the ones that get squirted in your nose.

[00:06:03] And that’s essentially a version of flu that can’t replicate enough to make you sick, but enough that your immune system can react. And those are basically the three kind of flavors of FDA approved vaccines for flu. The flu vaccine, the efficacy is actually pretty good against matched strains. But therein lies the issue, right. So we know flu season is going to happen. We know when it’s going to happen, right. It’s in the late fall or early winter. And because it takes a long time to produce the vaccines and formulate them and get them distributed to hospitals and physicians and companies like Walgreens and things where people get their flu shot. We basically have to start making those vaccines early, like well before flu season starts.

[00:06:50] And so essentially what people at the World Health Organization do is look at the viruses that are circulating and they predict which ones are going to then circulate when the next flu season comes along. So part of the reason that the flu vaccines don’t always fully protect people is that sometimes a different virus that we weren’t predicting circulates. So you get vaccinated with something that’s kind of close to the virus that you’re going to be exposed to, but not close enough to give you 100 percent protection. The other thing that can happen is sometimes, even if we picked correctly the viruses that we want to turn into vaccines, they don’t grow particularly well under vaccine production conditions. And so in those cases, we essentially select for viruses that grow better. So it’s feasible to produce these vaccines.

[00:07:42] And the viruses that grow that are slightly mutated relative to the viruses that are circulating in people. Any time you are vaccinated with something that’s not exactly what you’re being exposed to, the vaccine doesn’t work as well. But I will say that the vaccines, even if they’re not super efficacious in preventing infection, they still do a really great job of keeping you out of the hospital. So, I think that flu vaccines get a bad rap because it’s true: sometimes you can get your flu vaccine and then you can still get sick with the flu. The chances of that happening are decreased dramatically, but it can still happen. But basically, it almost always keeps you out of the hospital which is important as well.

Grace: [00:08:24] Yeah. And then talking more about these prevention efforts. So what we’ve seen this past year is that there was a really large decrease in flu cases during the pandemic. Can you talk a little bit about what may have caused that?

Nick: [00:08:38] Yeah, absolutely. So there’s a ton of surveillance. We call it surveillance for flu. Places all over the world are testing people when they come in and they’re sick or even just testing people off the street. We test them and we look for viruses. So we have a pretty good understanding of when flu is infecting people, what kinds of flu are infecting people. Over the last year, the last flu season, there was very little flu and nobody knows for sure, but almost certainly the answer is that wearing masks and social distancing helps prevent the spread of respiratory viruses. And flu is a respiratory virus, which transmits in the same way as SARS-CoV-2. So when you take into account that people already have immune responses to flu, everybody is exposed to a virus either by vaccination or infection. Essentially when they’re born, within the first couple of years, they have antibodies against flu.

Even if they haven’t been exposed to the exact strain, their immune system has seen something that’s similar. So, when you take people’s exposure histories, combined with getting flu vaccines, along with social distancing and mask wearing, essentially flu can’t circulate the human population. SARS-CoV-2 has also been controlled efficiently with behavioral practices, right. But the lack of pre-existing immunity is what flu benefits from. And SARS-CoV-2 has benefitted from that at least until recently, now that we have vaccines against it.

Grace: [00:10:08] And so do you think that these measures will be implemented every year for seasonal flu, or do you hope that they will or do you think we’ll probably just go back to getting the flu, getting sick and all that?

Nick: [00:10:20] Yeah. I mean, it’s an interesting debate. You know, it’s kind of been an experiment, right. We’ve never made everybody in the country wear masks and stay away from other people before. And now we know if we do that, we can stop transmission of these and other viruses, right.  The question is where’s the line? Right. Like we could say nobody’s going to be around anybody and everybody’s going to wear a bubble and keep them away. And for sure, that would stop the spread of infectious disease. I think that we’ve done a reasonably good job of controlling flu with vaccination. And the other thing that we have for flu that is really limited for SARS-CoV-2 are antivirals. These are pharmaceutical drugs that you can get when you go to the doctor and you’re sick with flu.

[00:11:05] So if you’re vaccinated and you have an exposure history, you’re already reasonably safe from flu. But if you still got sick and went to the hospital, we could give you drugs that will stop the virus. So we have a lot of tools in our arsenal to combat flu, which we didn’t have when the SARS-CoV-2 pandemic started. So this is the big difference. And that was the concern. That was really the impetus for the public health measures. Like if you get this thing, you’re on your own, right. There’s no medical interventions which are proven to be efficacious to stop you from dying from this disease. But now the vaccines against SARS-CoV-2 are amazing. And there are antivirals which are approved and tons that are in development. So I imagine that in the not too distant future, we’ll have more ways to fight this virus, which I think will help, so that we don’t need to wear masks for forever. But essentially at the end of the day, everybody will make that decision for themselves. But it is interesting to know kind of the magnitude of how much can be accomplished if you implemented those measures.

Grace: [00:12:07] Yeah. I don’t mind mask wearing. I had really bad allergies when I was growing up. And one of the things that they said you could do was wear a mask to prevent pollen from getting in your nose and mouth. And I was like, oh, but if I did that, everyone would think I was weird, you know, like really sick and avoid me. So for my own personal reasons, I hope that mask wearing is a little more accepted, at least in allergy season.

Nick: [00:12:34] Yeah. And it’s worth pointing out. We’re obviously talking about kind of a US centric look at these practices. But in other parts of the world and other cultures mask wearing is much more normal. That could be an outcome of this, right. It could be normalized in the United States and more of a culturally accepted practice. It’ll be interesting to see. Again, it’s kind of a big sociology experiment, how are people’s behaviors and actions being changed by a biological phenomenon like this.

Grace: [00:13:02] Yeah, certainly. So kind of moving into a discussion more about the SARS-CoV-2 pandemic. I really loved your lab’s bio on Twitter, which for our listeners was “we study influenza viruses that cause disease except for when we get interested in something else…” And I can only assume that that’s something else is SARS-CoV-2, given that you’ve recently been publishing a little bit in that space.

Nick: [00:13:27] Yeah, absolutely. So, the SARS-CoV-2 pandemic happened and at least for us, we’re working on these viruses that are kind of similar. Maybe we would have some insight, some things to add to the field. But more than that, the university shut down. Everybody went home. And the exception to that was if you were working on coronavirus. And so that helped increase our motivation to take on this new challenge because it was that or it was sit on the couch. There’s amazing groups who’ve been working on coronaviruses for a long time, and they’ve really led the effort, and a lot of this work. The viruses are different than influenza viruses, obviously, but they still use the same cell. They still use the same host, right.

[00:14:14] And one of the things that we had been working on with influenza was trying to understand what do they need to take over? What do they need to co-opt from the cell they’re infecting in order to replicate? We had been studying those questions for flu and we thought maybe that’s an area that we could work on to try and understand what the coronavirus would need to take from the cell such that it could efficiently replicate. I think people appreciate this, but there really is a dramatic difference in the genetic potential of a virus and its host. Even a big RNA virus, like coronavirus is a pretty big RNA virus, encodes less than 50 proteins, maybe on the order of 30 or so. Influenza viruses encode anywhere between 12 and, I don’t know, 20 proteins or something like this.

[00:14:58] And human cells encode genes is probably about 20,000. And if you take into account splice variants and these types of things it can be hundreds of thousands of different kinds of proteins that all have different jobs. So when you think the virus is going to be able to replicate itself, it’s come up with 20. It’s going to gather some from the host. And that’s an area of interest. A lot of groups, including ours. What does the virus need? Because it opens up kind of a practical option for not just stopping virus proteins, not just inhibiting virus proteins, but if you can inhibit either the interaction between a virus and a host factor that it needs or inhibit the factor that the viruses use. These are new possibilities for kind of antiviral treatments.

Grace: [00:15:38] Yeah. I mean, there was a lot of movement when the coronavirus pandemic started from different fields into virology. So I imagine you had a bit of a head start on that moving in from one field of virology into another compared to someone who moved from engineering, I don’t know something like that.

Nick: [00:15:56] Yeah. It was interesting, you know, science is always better with diverse perspectives. And I think the field has benefited from somebody who thinks about a totally different question or usually thinks about different questions and now says based on how I think about things, what do I think is going on? It has moved the field forward rapidly. And, we know a lot more about these viruses than we did a year ago. I mean, that’s been the good side of the spirit of scientific collaboration and discovery that’s been all focused in this area.

Grace: [00:16:29] How do you think that the pandemic has changed the way that the general public thinks about virology?

Nick: [00:16:36] Yeah. I mean, that’s an interesting question. I think people think about viruses now. For some people the level of resolution is germs. And there are things that make you sick. There’s all kinds of things, right. There’s bacteria, there’s viruses, there’s protests and all kinds of things. The idea of a virus has certainly come to the forefront in people’s mind. I mean, which is cool, right. These are things that we think about all the time. But not everybody does. So it’s kind of cool to have more universal recognition of the types of questions and things that we’re interested in. I think there’s also been an interesting kind of realization and attention paid to the scientific process in general, which I think is really going to be a helpful thing to come out of the pandemic because it costs money, right.

[00:17:24] A lot of the biomedical research in the US is funded by taxpayers. I mean, the lion’s share for sure, is people’s taxes, right. And the question is, what are we studying and how is it helpful? What’s the return on this? Right. And, you know, I would say that the coronavirus pandemic has demonstrated what this return is. You know, it sometimes frustrates people that we don’t have all the answers right at the beginning. But when there’s a question, we can activate this biomedical research machine. We can understand things like how transmissible are these viruses? How long do they stay in the air? Can you transmit them by touching? These are all experiments that are done that answer those questions.

[00:18:03] And I mean, the most dramatic one is the development of these vaccines. I mean, the development of vaccines takes decades. And the fact that collaborative teams were able to come together, pharmaceutical companies and industry and academic institutions for testing and development of all these things, and have a vaccine essentially a year that is highly efficacious. I mean, this is what that money is going to. Discoveries that enable this type of stuff. This is the payoff for supporting that kind of endeavor.

Grace: [00:18:31] Do you think that sort of interest will continue beyond the pandemic or do you think people will as soon as it’s over, be like, oh, forget biology, let’s go back to normal life?

Nick: [00:18:43] Yeah. I mean, I think that’s a natural tendency that will occur to some extent. But, hopefully this will remain fresh enough in people’s minds that I mean, you see these things in congressional hearings, right. Where they’re discussing the budget on science. Why do we need to give this much money to research? At least for the short term and hopefully for longer we’ll be able to say this is why. God forbid, there’s another pandemic. But if there is, we need these people working on these kinds of questions and developing these things. We can’t be caught flat footed.

Grace: [00:19:11] And would you care to make any predictions about what you think the next pandemic might be? Do you think it’ll be pandemic flu or another coronavirus or Ebola? What do you think?

Nick: [00:19:22] Yeah, historically, at least from 1900 on, there’s been a series of pandemics and they’ve all been flu pandemics. There’s been coronavirus outbreaks and they’ve been epidemics. This is like SARS and MERS, the viruses that are very similar to SARS-CoV-2, but they were regionally contained. So just numbers wise, starting again with the 1918 Spanish influenza epidemic or pandemic. We’ve had a number of these flu. Most recently in 2009, the H1N1 swine flu pandemic. They’ve all been flu. So I think if you were betting, you would bet on it being a flu pandemic, the next one. But you know, who knows? And as the environment changes and people are living more places in more proximity to kind of reservoirs of animals. I mean, we’re just increasing the contact between people and things they weren’t exposed to before, including viruses.

[00:20:18] You know, I think one thing that will come out of it is the flu field has been engaged in the surveillance that I was talking about earlier in the development of interventions or countermeasures for pandemics. With the idea that another one is going to happen at some point, we need to be prepared. There is much, much less emphasis on coronaviruses because we hadn’t had a coronavirus pandemic at least since the molecular age. And we could really identify what these viruses were that were causing disease. But obviously, now it’s clear that they can. So I think the same type of effort and in terms of surveillance and predicting these things will be applied to coronavirus. And if it should happen again, I think that we will be much better prepared.

Grace: [00:20:56] Yeah. And I know there’s been some talk in the science community that the frequency of pandemics might be increasing. Can you speak a little bit on that and why you think that may be?

Nick: [00:21:08] Right. So it’s impossible to predict. But if you look at the things that affect transmission of disease, which is required for a pandemic, people move across the world at an unprecedented rate compared to what we used to do. That I can be anywhere in the world and can be anywhere else in the world within 36 or 48 hours is a bad thing in terms of transmitting pathogens around. So that’s a big part of it. The second thing is that we have more people, right. More people closer together. The cities are bigger and it’s much easier for pathogens to transmit when there’s just more people that are infectable around. So that’s a factor. And then, of course, this is what I was referring to before: a lot of viruses are transmitted by vectors, mosquitoes, for example, right. And as climate changes, animals and insects change their distribution as well. So populations come into contact with these viruses or the animals or the insects that carry viruses, which then facilitates these spillover events, which probably happened with SARS-CoV-2 at some point.

Grace: [00:22:13] Do you think there’s anything that we can do to reduce the rate of these pandemics happen?

Nick: [00:22:18] I mean, these are hard questions, right. Because people will have to live and work and travel where they travel and these types of things. Obviously dramatic things like that. I don’t think that’s likely to happen. I think that what’s much more likely in terms of being able to prevent pandemics or to contain them as epidemics and prevent them from becoming pandemics is just surveillance and countermeasures, right. If we know when these things start, we can detect them right away and quarantine people or if we can rapidly make countermeasures, right. If we could have made these vaccines within a month of detecting the virus, that would have presumably made a big impact on how far the virus was able to spread. So I think that’s probably where most of the effort is going to be focused. Understanding where these things are coming from. Understand quickly when they’re in the human population, and then being able to respond rapidly is probably the direction to go.

Grace: [00:23:15] So moving into more of a discussion about you as a person and as a scientist, I’d love to hear a little bit about how you came to virology and influenza and how you made it your way to Duke? Yeah. So start from whenever you first got interested in science at all.

Nick: [00:23:34] Yeah, it’s been a while. So I guess it started in high school. When I was in high school, I took all of the science classes that were offered and there was nothing else to take. And then I ended up going to the university. It was in my town, Utah State University, for credit to work in a research lab. And then when I went to college, when I was doing my undergraduate studies, I had this experience now like I was qualified or at least more qualified to work in a laboratory. And so I got a job working in a lab to pay the bills. I was in a great lab, they gave me some opportunities to develop some of my scientific skills. And then basically I became good at it. You know, I thought this is something that I like doing and something that I think I can do well.

[00:24:19] And from there, I went to graduate school based on that interest. My undergraduate degree is actually in bacteriology. And I went to graduate school to work on bacteria. And then in this kind of fortuitous event, the guy that I ended up doing my thesis work with, we were in the elevator. He was asking me how graduate school is going. We start graduate school. You do rotations where you essentially work in a few different labs and kind of figure out what’s a good fit for you. He asked how my rotations were going, and I said I didn’t know which lab I was going to rotate in next. He said, well, we just got this grant to work on a virus called dengue virus.

[00:24:52] I don’t know anything about dengue virus, but it sounded kind of exotic and cool. And so he said, come rotate if you want. And I ended up working on that virus. It’s one of those mosquito transmitted viruses that’s prevalent in the tropics. And then I was good at virus research. That’s what I was qualified to do. So I kind of kept doing it. I did my postdoc on influenza, then I’ve been working on flu and the respiratory viruses ever since.

Grace: [00:25:18] I always love hearing about the serendipity of people’s journey. I think that’s really awesome. What exactly was your thesis research on?

Nick: [00:25:28] Yeah. So we were working on dengue virus. And one of the questions that we were really interested in was the interactions between the virus and the hosts. And the reason that we were interested in that question in particular was the virus replicates in mosquitoes, right. It lives in mosquitoes. And that’s how you get it. A mosquito bites you. And so this virus has to exist in mosquito cells and in human cells. And those are really different environments, right. A mosquito is very different from human. And so, anyway, those were some of the questions that we were interested in. We ended up doing a series of experiments looking at the role of cholesterol and fatty acids, which make up membranes and how the virus essentially reprogrammed the host cell to make the membranes that it needs for its replication and its assembly.

[00:26:13] That’s what my thesis was on. We published a couple papers on that. Towards the end of it. So we had looked at the host side, and what we really wanted to then do was look at the virus side of things. But the genetic tools for dengue virus at that time were not particularly developed. It’s really hard to make a mutant virus. We were trying to make a virus that couldn’t reprogram the host cell to move these membranes around and see what would really happen. So anyway, this was something we were interested in but were unable to do at the time. And then I was thinking about what I wanted to work on next and this is one of the reasons that I picked flu, because flu had a really good genetic system and there were a lot of things you could do with the virus. And that’s kind of what I was interested in getting trained to learn how to do.

Grace: [00:26:54] So when you aren’t busy being a scientist, what do you do? What do you do for fun? Who’s Nick the nonscientist?

Nick: [00:27:03] Well as you know the process of science takes a lot of time. My wife works in the lab with me, so it’s kind of a family business. And we have two young sons. So between the two of us working and then taking care of the kids, that’s essentially the full schedule of events for us.

Grace: [00:27:20] That’s the life. That’s awesome. What’s it like to work with your spouse?

Nick: [00:27:24] Yeah. As you can imagine there’s pros and cons. There’s way more pros. I think in our particular case, we met in graduate school. And so we’ve always kind of interacted doing science and talking about science and troubleshooting experiments. And so it’s just kind of been a natural evolution. And now it’s really great because, you know, she gets it right. Like when there’s an experiment that has to be done or a time point that’s really late or something spills into the weekend or something like that, it doesn’t take explanation. It’s just as things go and she knows the details, she gets it and we figure out how to make it work. That’s been a huge benefit.

Grace: [00:27:59] It’s better than someone who has no experience in science, because it really does take a lot of work and odd hours and that’s pretty cool. And you always have a carpool, buddy.

Nick: [00:28:09] Yeah, that’s right. Reduce our carbon footprint.

Grace: [00:28:13] Yeah. So as we wrap up the episode. What sorts of advice would you give to people who might be entering the field of virology today?

Nick: [00:28:24] I guess I would have two pieces of advice. The first is that you really need to become good. And what I mean by that is there are metrics in science, right. If you publish papers or you get fellowships or you get grants or these kinds of things. And I think a lot of people focus on hitting tangible metrics so you can put a line on your resume. I got this or I published this or whatever. And I think sometimes there’s less of an emphasis on really becoming an expert in the process of doing science. How do you set up the best experiment so you can make the clearest conclusion from it? And so that’s what I would tell people to start with. All the rest of that stuff comes if you’re doing good, thorough, reproducible science.

[00:29:09] You get all the rest of that, but at the beginning, that’s really what the focus should be on. And it’s satisfying. You know, sometimes from a CV building point of view because there’s nothing you can see, right. You’re thinking about questions better and more precisely. That’s really important. The other thing I would say is that there’s lots of reasons why people go into different fields or pick different topics to study. And I think that science is an endeavor where it’s really important to be excited about the specific thing that you’re working on. And it means different things for different people, even within the context of like working on one virus. There’s all kinds of different directions that you can approach it from. And figuring out what excites you, which little nuance of the questions are you most excited about?

[00:29:55] That’s no small part of being able to be successful and invest what’s required and when an experiment doesn’t work, something like that. I think it’s that that gets you out of bed the next morning. You know, you really care about it and you’re really working on the right question, helps you get through the lows, which always happen.

Grace: [00:30:11] Two excellent pieces of advice. And I hope our listeners can take those away with them after this episode. Thank you so much for joining me, Nick. It was really awesome talking to you and hearing your perspective about viruses and the pandemics and advice for life.

Nick: [00:30:27] Thank you. It’s been a pleasure.

Transcript of Episode 38: Stacy Horner

Grace: [00:00:00] Welcome to The Bioinformatics CRO podcast. My name is Grace Ratley, and today I’m joined by Dr. Stacy Horner. Stacy is co-director for the Duke University School of Medicine, Center for RNA Biology, as well as associate professor. Welcome, Stacy.

Stacy: [00:00:15] Thank you for having me, Grace.

Grace [00:00:16] I’m excited to have you on the podcast. So can you tell us a little bit about your research on flaviviruses?

Stacy: [00:00:22] Yeah. So my lab studies virus-host interactions. And so what that means is that we’re really interested in how viruses infect cells, how cells sense these viral infections, and then how the cells in our body fight back to viruses and then how viruses try to get around that immune response. And so, as you mentioned, we do this in the context of viruses in the flaviviridae family. So that includes viruses like hepatitis C virus, dengue virus, Zika virus, West Nile virus, all viruses that many of you may have heard of before.

[00:00:53] And I mentioned that these are positive strand RNA viruses. So they’re very similar in some respects to SARS coronavirus too, which is the virus that’s causing the current COVID-19 outbreak, which also has a positive-sense RNA genome. And so fundamentally, what we’re learning and how we study how these flaviviridae viruses interact with the cells in our body, our principles that can actually go beyond just these viruses, because many RNA viruses are viruses that generally interact with cells in the same way.

Grace: [00:01:23] Can you tell us a little bit about hepatitis C in particular? So hepatitis C does have an effective treatment. So why is it important for us to continue studying this virus?

Stacy: [00:01:35] Yeah. So hepatitis C virus, as you know, we have great direct acting antivirals to this virus that can actually cure the virus. So unlike HIV, which when you’re infected, it goes into kind of like a latent state and heads out of your body, we can eliminate HCV from people. So that’s very exciting. And in fact, the scientists who discovered aspects of hepatitis C virus infection and how to cure it, actually just won the Nobel prize last year. And so the reason why it’s important to study this virus, even though it’s already cured, are two big reasons.

[00:02:09] The first is that there’s actually no vaccine, so we can’t prevent infection. And one of the largest sources of new HCV infections every year is through IV drug users. And in fact, at least a year or two ago, the number of new infections per day was equivalent to the number of people being cured from the virus. And so the only way to actually eliminate the virus is to prevent infection. And we really need a vaccine for that. And so while my lab isn’t studying the immune responses that would lead to a functional vaccine, there are many labs that are doing this. And I think having a vaccine to prevent infection by HIV is something that we really need.

[00:02:46] So that’s one reason why people in general should still study it. But from my lab, we like to study this virus because it interacts with our host cells in a lot of interesting ways. And by understanding how this virus interacts with our host cells or causes infection, we could actually learn general principles about how viruses can cause disease. And so we can use this virus as a model virus to learn how these things happen and then compare and contrast to other viruses to see if it’s similar or different.

Grace: [00:03:15] And you mentioned that some of the RNA biology is similar in both flaviviruses as well as SARS-CoV2. Can you talk a little bit about that and what you’ve found so far?

Stacy: [00:03:28] Yeah. So the most important thing that’s similar is that their genomes are what we call a positive sense polarity. And so that means when the virus infects a cell, the viral capsid un-coats, the viral RNA gets into the cell and immediately that RNA is translated to make the viral proteins for both viruses, the flaviviridae and for coronaviruses in general. The first thing that needs to happen is that the viral RNA dependent RNA polymerase gets made and this viral protein can then replicate the viral RNA to make more copies. Actually, in the viral RNA dependent RNA polymerase is a big drug target for hepatitis C virus at some of our drugs that are in the clinic are used to that.

[00:04:10] And you could imagine that for coronaviruses it would be a similar type of idea. One major difference between the two viruses and this is why actually there isn’t a drug that targets the coronavirus, RNA dependent RNA polymerases is that the coronavirus because its genome is very large. So it’s RNA genome is 30 kilobases long, whereas viruses in the flaviviridae family are around 10 kilobases long. So it’s three times bigger. And so because of that, when the viral polymerase copies the genome, it makes many errors, right. And the longer genome, the more areas that you can make. And so coronaviruses actually encode a proofreading enzyme that can fix mistakes.

[00:04:52] Many times in virology, we want to target the polymerase, but actually in coronaviruses it could maybe fix mistakes that happen because they also include what we call exonuclease that can help to repair the genome. So I think the exonuclease might be a good molecule to target in coronaviruses. And in fact, a new paper came out, I think today sharing the crystal structure of that protein of the virus. So that’s very exciting for people who study coronaviruses.

Grace: [00:05:20] So how do you think the pandemic has changed the way that people think about the field of virology as a whole?

Stacy: [00:05:29] Right. So I think it depends on who you ask, as we all know. So in my family, I’m very popular right now because a lot of my family has a lot of questions like probably many of us do about: how the virus works, how do you get it, what’s safe, what’s not. And the fun part of being a virologist is that I actually read the scientific papers. And so I know the data in the papers and what seems right and what’s not. And one of the interesting things in the last year is that the media has been covering the virus a lot and sometimes they aren’t able to quite capture the nuance of science, right.

[00:06:08] And so sometimes I look at a paper and I come to a conclusion and then I read a news article, and they come through a different conclusion. So that’s something that’s been pretty interesting. And for my family, as we’re trying to all make decisions to keep us safe, I’ve actually relayed some of like, well, I know that’s what the article said, but this is kind of what I think based on the data in this paper. And so if we go to a broad level now, a lot of people are, what we call armchair virologists, which is kind of exciting to have new people from other fields interested in how viruses work.

[00:06:40] We’ve made a lot of progress in only about a year, which is really remarkable, right. We didn’t even know the virus existed and now we have a vaccine that’s in like at least half the people over 18 in the United States, right. So I think that’s very exciting. One of the hard things, one of the challenges has been that many people think they know everything about virology and they come in from a different field. And there’s actually a pretty well established coronavirus literature. And so I know some of my colleagues who are coronavirologists have been a little bit irritated if people don’t know the literature.

[00:07:13] So that’s something like the scientific sphere. But if we go to a global sphere, a lot of people say, well, I am really good at predicting these things. I’m really good at epidemiology. I am really smart generally. And so I want to tell you how you can cure this virus. I think we can all appreciate everyone’s efforts to come together on this. Sometimes we wish that people would spend a little bit more time preparing and actually going to the experts who really do know what’s happening, the advantage of people from outside as we get more creative solutions, the disadvantage is they have to get up to speed. And as someone who’s been studying viruses for 20 years, that’s a lot of speed to catch up to, right.

Grace: [00:07:51] Yes, certainly. We had a person on the podcast recently who was using CRISPR-based technologies for COVID diagnostics, which was a really exciting new development. They’ve been using it to try and diagnose other conditions. And when the pandemic started, they quickly jumped in and helped out there. Can you think of some of the more exciting technological advances that have come as a result of people entering the field of virology from other fields?

Stacy: [00:08:20] Yeah. So I think, one, that is pretty exciting, and I don’t actually know the people who did the work, but I’m sure you can find the publication for the listeners, is novel methods to detect coronavirus infection. So currently you have to do like a PCR test, which takes a while and is expensive. You could do an antigen test which is much cheaper. But I read a report of a study that can take the air that people are breathing and then do mass spec on the air–I think it’s mass spectrometry–to identify the exact molecular weights of the things that are in that air and they could actually detect the virus.

[00:08:58] This is super cheap. You could put it in like train stations. I thought that was really cool. So that’s like some of the really cool technology that you would have expected. And I think a lot of this comes from, you know, maybe people in the engineering world. And then, you know the other big thing for my field that I think is very exciting is the fact that the COVID vaccine is the first mRNA based vaccine that has gone into people and been approved for this kind of global use as emergency authorized, right. So while a lot of the technology for mRNA vaccines has been around for a really long time, it wasn’t until right now that we had to make a vaccine really fast, what are we going to do?

[00:09:39] They kind of had done all the research over the years to be ready right for this moment. And the fact that it works so well and so quickly is very exciting, because that means that then we can use this technology for future pandemics, for a lot of other diseases or viral infections. I would say from my point of view, the fact that you could go from not knowing a virus to get an FDA approved vaccine in people in a year… I actually did not think it was possible, and it actually was, which is really exciting.

Grace: [00:10:07] Yeah, it is. It’s so exciting. So one of the things that I’ve always wondered about is why did it take us like a pandemic to get these things to market? Like what do you think is the difference?

Stacy: [00:10:18] Yeah. So it was really an accelerated timeline, a lot of resources put into it in an urgent need. So if you talk to people who work for companies and do clinical trials, clinical trials are very expensive. And you start with your phase one, which is a small safety trial, then you start the phase two, which is like a little more safety, a little more efficacy. And then you go to phase three and these will often be spread apart by several years as they review the data and think about as a company, what strategy do you want to go after. Can we raise the funds for the Phase three trial, which is very expensive?

[00:10:52] And so in this particular case, they just did all those back to back to back. They didn’t do a lot of the optimization that goes through, like, what is the right dosing? There was a little bit of that, but not as much as is normally done. And I just had to say, let’s go for it. And so because the vaccines had the support from the federal government, they could actually take the risk to go through all the trials, right. Because they’re very expensive. If you fail, you just kind of lose all that money. But because the government partnered or was willing to buy the doses, that really took that risk out of the picture for those companies.

[00:11:27] So I think that was the big thing. And also that the other big thing that I would say like from a virology point of view is that the pandemic was still ongoing. And so you could actually do the clinical trials. You might remember a few years ago in 2016, there was Zika virus pandemic. And in fact there were Zika virus vaccine candidates, even a mRNA vaccine candidate. But the Zika pandemic pretty much died out. And so you had a vaccine, but you had no way to test its efficacy. And so in this particular case, we are still testing a lot of efficacy. And so you need enough study participants to do the trial. You have to recruit them.

[00:12:04] And in this case, we had so many people that you could recruit for a study. It was actually able to get your end points, you know which is how effective is it in a controlled study with a matched-control group. So I think that was very exciting. And some of the reasons we were able to accelerate are support from the government, financial backing, the urgent need, and having a huge study population.

Grace: [00:12:25] So do you think there will be long lasting changes in the way that the government interacts with and funds virology research or medical research in general?

Stacy: [00:12:35] We would hope so. Others might know that as a professor at a medical school, one of the things we have to do is raise grant money, right. Grants fund our research. So the school doesn’t give us the money for that. They support us, but we have to get the money from the government. My dad always says this should make getting grants easier for you. You wish it would be the case. I think the NIH has always had a commitment to funding virology, but there are other diseases that are really important, too. And so it’s really kind of the budget isn’t large enough for the research that we really need.

[00:13:09] And so I’ve heard from people in Congress and many senators are supportive of biomedical research and increasing funds, but there’s just not enough money to go around. I think that one of the biggest lessons we learned this year is funding basic science is important. So the technology that led to the mRNA vaccines came from basic science and a key discovery that was made in 2004-2005. And we didn’t realize how important that was? I don’t want to predict, but I imagine they’ll get a Nobel Prize for it. Actually, that would be my prediction.

[00:13:45] So basic science is important. All science is important. And fundamentally, we need to just increase our funding for all science. That’s my take on it. Hopefully a lot of people have realized the importance of funding basic science from this pandemic. So the other thing that I want to say about funding science is I think that this pandemic has really also shared with us the lack of a public health infrastructure in the United States. We used to have a strong public health infrastructure. But in this case, when a new virus came out, we should have known what to do. We should have had a plan and there really wasn’t.

[00:14:21] And that’s because things like epidemiology, your local county small town public health departments have been underfunded a lot and that’s over the last 10-15 years. And so I think you need to keep those things funded because every day you might not need them, but when you need them, you really need them. People like to listen to people in their communities rather than the federal government. And so funding these kind of pandemic preparedness at the local level I also think is really important.

Grace: [00:14:49] Yeah. So I was a public health major in college. And one of the things I was super surprised to find was that if there was some sort of like Ebola outbreak in a particular place, the people in charge would be the local public health officials. Like it wouldn’t be the federal government coming in. It’s not like the FBI takes over a criminal case or something. It’s those local governments leading that response. So, yeah, I would have to definitely agree with you that investing in those infrastructures is very important. So if you were to put yourself in the shoes of the people who are deciding where funding goes, if you had maybe a million dollars to give, what sorts of research efforts would you be most excited to invest in?

Stacy: [00:15:30] So, first of all, I would say that a million dollars is enough to fund three people for five years. So we actually need to go like one hundred million dollars. But I think one of the biggest problems actually in our biomedical scientific enterprise is the lack of diversity in science. And so this means black and brown or people of color really need to be better funded. We know that there is data that the grant peer review process is actually quite biased in a number of different ways. And we also know that diverse teams are more successful.

[00:16:06] And so not only from an ethical point of view, but from actually how are we going to solve problems as a country, as a world? How do we fight the next pandemic? We need people from every walk of life to see scientists that look like them, to want to become scientists and to bring their creativity, their energy and their communities to the table. And so if I had a lot of money, I would use all of the existing research that has been done by a lot of my colleagues around the country into this space and change policy to fund black and brown scientists.

Grace: [00:16:39] What sorts of policies would you change? Would it be maybe replacing the people who are deciding where the money goes or would it be educating them or what kinds of changes are needed to drive that increase in diversity?

Stacy: [00:16:54] You know, this is not my area of expertise, so I’m not going to presume to know the right answers. But I think there has to be a real commitment from the NIH to not just talk about changing these problems, but to do bold solutions. And so they have advisory committees with the right people on them who can help them make these decisions. I think the peer reviewers of the grants are people like me who are trained in such a way to value certain things in the grant review process that will disadvantage certain groups.

[00:17:24] We all know this. So it needs to be led by people on the review panel saying, like, hey, those words that you use, they’re actually not okay. That’s not what we’re looking for, we are looking for good science. And that’s kind of risky. You could also just commit to funding more people of color. Everyone will say like, oh, well, we want the science to be a certain quality. Trust me, the quality is good. It’s not a problem of that. It’s about the structures in place are biased in judging what is good science or not good science. And so I think that’s a problem.

[00:17:58] At an institutional level, we need to hire more people of color. But not only do we need to hire them, we need to make them feel supportive, give them mentoring, treat them as colleagues, not as people who can make our quota of having more people in the pictures. And so access to mentoring, not overburdening them with service and thinking about our promotion criteria. So someone like me, I just got tenure. So I’m now an associate professor, but I went through that process. That whole process is biased and favors certain things that will be disadvantaged by other people who do not fit into the mold of what a scientist or what publishing should look like. Many of us know what’s wrong. It’s pretty bold move that to change that system. And I think that’s what really needs to happen.

Grace: [00:18:48] Do you think that sort of change would take a long time or do you think it would need to be something like kind of like what we saw in the last year with the Black Lives Matter movement, where it’s just like intense social support moving towards a single goal?

Stacy: [00:19:01] Yeah. So I think some things are easy and some things are harder. You’re not going to change everyone’s mind. That’s not why you do it right. I think those kinds of things could be harder. But there are easy things you could do. For example, when people are coming up for tenure, have them talk about their commitment to diversity, equity and inclusion. Make white people do some of that work, too. That’s like something that actually costs zero money. It takes time, but I think that we should be giving that time. So, yes, mentoring programs, that don’t actually cost that much money.

[00:19:32] Fundamentally, I think we should just hire more black people or brown people as assistant professors, but then also support them. And so that means every institution needs to take a hard look at how they have supported people like this. What do they do to help them get tenure? Because processes like promotion or even if you’re a PhD student, graduation, those are all biased and they’re geared towards a white male demographic, which is kind of historically how it was. But that doesn’t mean that it’s right and that doesn’t mean that we should be still be doing that.

Grace: [00:20:05] So kind of changing gears a little bit. I’d like to get into what inspired you to become a scientist. And when you were growing up, were you always really interested in science?

Stacy: [00:20:16] So when I was a child, I was always interested in how things work. I remember my mom had a book, How Things Work. I just was like that’s cool. I want to learn a little bit more about that. I always hated science class in school because it just seemed really boring and like memorization, so I actually never wanted to be a scientist per se from a young age, but I was curious about how things worked. I have a very defining moment, which is when I don’t know what year it came out, but it has to do with the movie Jurassic Park.

[00:20:48] I think I was probably a junior in high school somewhere around that time. I remember riding the car with my dad and my dad liked to listen to NPR. And NPR was playing a show about the movie Jurassic Park. So I hadn’t read the book yet, but I was listening to this podcast, not podcasts, this radio show about Jurassic Park. And they were talking about the whole premise of Jurassic Park is that these people, we can recreate dinosaurs from fossilized mosquitoes. You could extract the DNA–because some mosquito bit the dinosaur–pull out the dinosaur DNA and then do some kind of magic, basically to make a dinosaur.

[00:21:32] Well, when I heard about the science behind that, there must have been some scientists talking about it, I just thought that was the coolest thing that I’d ever heard. You could go from the DNA to making a whole animal or an organism. So I went back to my high school teacher and I asked him about DNA. I guess I missed that day in biology class because I was like I want to study DNA. What is that field called? And he said biochemistry, chapter twenty four. I opened the book and it had a picture of DNA. And I was like, okay I’m going to be a biochemistry major. It really was as simple as that.

[00:22:05] And probably as naive as that, to be honest. From then on that’s what I wanted to do. I pretty much didn’t know what that meant. I didn’t know what being a scientist would be like. One of my grandpa’s was a chemistry professor at a small school in Minnesota. So in theory, I should have understood what that meant. But I thought maybe I’ll do pre-med biochemistry. When I got into college, I realized I didn’t want to do pre-med because I was really interested in the small little things like nucleic acids. And so I didn’t want to study bone or the body. I wanted to get right into those details. I’m like, how do they work? Like, how do molecules find each other? That sounds really cool.

[00:22:45] So that was why I wanted to be a scientist. The reason why I wanted to be a virologist is not as clear to me. Although I do know that my first or second year in college, I’m from Minnesota. I went to a small liberal arts college there, called this Gustavus Adolphus College. And every year at Gustavus, they hold something called the Nobel Conference. Gustavus is a Swedish Lutheran college. I’m not Swedish, but that’s the school I went to. And so they have an affiliation with the Nobel Foundation in Sweden that helps them organize a really large conference for a very small liberal arts college in middle-of-nowhere, Minnesota.

[00:23:22] And this particular conference was on viruses. So they brought in all these world leading virologists to the conference to talk about virology. And I’m assuming that that’s what really piqued my interest. Because after that point, I knew I wanted to be a virologist. And before that, I don’t have any memory of that. So why did I want to be a virologist? It was cool to me that viruses can go on to cells, change cell biology. And not only help you learn about the viruses, but also help you learn about the cells that they infect. So a lot of key discoveries in biology had been made by using viruses as tools to study this biology. I thought that was really cool and that’s why I wanted to do it.

Grace: [00:24:02] Yeah, I totally get that. Viruses are just little machines and are very interesting. So when you were pursuing virology, were there any moments where you were like, this isn’t at all what I expected? Or what were your expectations versus reality when it came to being a scientist?

Stacy: [00:24:21] So I did my Ph.D. at Yale. And when I was a graduate student at Yale, I, like many graduate students, struggled at some level. Mostly because I felt like I wasn’t good enough to be a scientist or to be a professor. From a young age I actually always wanted to be a teacher, and so being a professor to me seemed like the career path that I would want to choose. And I remember going to seminars or going to research and progress talks about other graduate students and thinking, wow, their science is so much cooler than mine. They are so much smarter than I am.

[00:24:59] And then you hear all about how like it’s really hard to get a grant. It’s really hard to be a professor. And I thought, well, you know, I’m not smarter than these five people I know. So I’ll never be a professor. I’ll never get the grant. And so that was really disheartening. You could do all this work and then try to get a job as a professor, you would never get one. And that seemed like a little too risky for me. It made sense to finish the PhD, and for all students out there, I would say if you can stick it out, it makes sense to stick it out.

[00:25:29] There are obviously cases where that doesn’t make sense. And so I’m not going to speak to all situations, but having a PhD opens a lot of doors. And when I thought about what I really wanted to do, I thought that being a virus hunter sounds cool. I want to travel the world and do that. And so I was already working on my application to try to get into the Johns Hopkins public health program for people with a PhD And then maybe work for the CDC or the WHO as a virus hunter. So I was putting that application together. Now I’m really glad I didn’t did that because it actually is like a lot less glamorous than you see it in the movies.

[00:26:04] I was talking to some of my committee members and then to my graduate advisor who really convinced me that I had what it takes to be a scientist and to be a professor, which is something I hadn’t seen in myself. I just thought I was so different than all the people I had seen in a science professors. I was so different than my classmates who were like, “I want to be a PI.” I don’t want to do that. I don’t look like you. I don’t do the same things you do. It didn’t seem like it was something that I could do or that I could be part of that community.

[00:26:38] Until my graduate advisor really stepped up and said, “Stacy, you have what it takes.” And also one of my committee members and so I’ll name them here. So Daniel DiMaio at Yale was my PhD advisor, and he was the one who really believed in me, as well as one of my committee members, Peter Lengyel, who’s now passed away. And so the two of them said, try postdocs, see if you like it, do one year. And if you don’t like it, I’ll write you a letter wherever you want. But we think you have what it takes. Actually after that it was very clear to me that I wanted to go and do a postdoc and then keep going on in science on the academic track.

[00:27:14] Because actually I think science is cool. I’m curious. I like thinking about big problems. I still didn’t know if I was creative enough to be like a cool, innovative scientist. But one thing that I say to all students is that by about year four of my postdoc after my PhD, I realized that I was creative. I had a lot of good ideas and so, you know, college student, first year graduate student, first year postdoc: you don’t have to know if you have what it takes. You just need to be curious and want to try. You don’t have all the information, I think, to make the decision to pull yourself out of the running.

[00:27:48] Now, if you don’t want to do that, that’s cool. But if you think you might want to do that, then try to find a mentor who can really talk to you about what the whole process was like for them. Because a lot of people have a story somewhat similar to mine, especially with women. We often have very different stories, unlike the kind of classic path. Everyone who thinks that they might want to do it owes it to themselves to try to find a mentor who can help encourage them along the way.

Grace: [00:28:13] So what do you think are the best qualities of a mentor? How can mentors better support their students to instil confidence in them or maybe a realistic view of themselves?

Stacy: [00:28:25] So I think if you’re a trainee, you should not think about only having one mentor but having multiple mentors. And these mentors will help you in all aspects of your career. And ideally you want to feel comfortable enough talking to them and being honest with them because they can only help you if they know what’s going on inside that brain of yours. And so you might have mentors that help you. For example, for me, that could be a mentor who looks like me. So that can be helpful. It can be helpful to have mentor who’s in your research area that you’re interested in.

[00:28:59] It can also be important to have a mentor that has a career like you have. So those are all things that one can look for in what we might call a mentoring team. Now things that mentors can do to help trainees as well–I actually just took six hours of mentor training and we learned a lot of thing. But I would say the biggest problem between a mentor and a mentee is ineffective communication. And so as a mentor understanding, that communicating with your mentee is important, the words that you use matter, and that our job as a mentor is not to get them to be you, but to help them assess their skills and figure out what they want to do based on their skills. And so I think those are effective mentors.

[00:29:43] And for mentees I mean, be open, be able to communicate and also tell your mentor what you need. As a mentor I really want my mentees or people that I interact with, even in class, to succeed. But I can’t read your mind. So it’s an important thing for the mentees to tell their mentor how they’re feeling. And for mentors to learn skills, to bring out the things that they need to help them be effective mentors.

Grace: [00:30:08] So you’ve been at a few different universities. So you’ve been at a small liberal arts college. You went to Yale for graduate school and then you moved to Duke. So what are some of the biggest differences that you’ve seen in the different cultures between those universities and some of the things that you like about being a Duke?

Stacy: [00:30:25] I think Duke and Yale are very similar in a lot of respects, to be quite honest. I did a postdoc at University of Washington in Seattle and that was very different as a public school. So that’s bigger, less personalized. And then a big difference between Duke and Yale is that Yale is a much older institution than Duke is. So fundamentally, there are some differences related to that. But I like being at Duke. And one of the reasons I also liked being at Yale is that they’re a little bit smaller. And so this is also related to kind of my undergrad, which was small liberal arts college.

[00:31:00] It’s so much easier, I think as a faculty and probably as a student to find different types of communities and interface beyond just your department into like the whole community. So I could go to like an English lecture just because I thought that was interesting. So I really like this kind of smaller aspect. What I also really like about Duke is how collaborative and interdisciplinary it is. It is quite common for me to be talking to someone in the biochemistry department, the chemistry department, or immunology.

[00:31:35] And I think this is really exciting. When I first came to Duke eight years ago or something like that, there were actually a lot of new assistant professors that were hired at that time. And so as a newer professor, I had all these friends. I didn’t know that I would be friends with other professors, but I certainly am. And so that’s fun because we all go through the same career challenges. But then, like, inevitably we get to talking about science or some new technique. Then you can build like, real collaborations and do cool discoveries together.

[00:32:08] So I think at a place like Duke it’s pretty easy. The pandemic has changed a lot of that. So for the new assistant professors who started in the last year or so, this is going to be one of the big challenges, helping them find that community in the next couple of years as they kind of do what I already did.

Grace: [00:32:24] Do you think a lot of the remote aspects of doing work, like maybe working from home a couple of times a week will be maintained?

Stacy: [00:32:30] I mean, I think some of them well. For example, certain kinds of meetings are really easy to have remotely. Everyone’s really comfortable with that. Having these Zoom meetings with colleagues all across the country a few years ago, I mean people did it, but it wasn’t super popular. Now, I think it would be pretty easy to move school-wide things to Zoom fairly easily like a training of some kind. But I do think that the real personal connection or the things that happen when you’re getting coffee, the people you run into, who you just casually talk with, those things will still be important in the future.

[00:33:09] I think one other thing that the pandemic has really taught us is the importance of having structures to promote or support people who have children. You actually can’t work from home if you also have children at home for school. It’s just not possible. And so I think those kinds of challenges and how we support folks with children, you know, pre-tenure with their grants at the university level. I think universities really need to support these things because they have caused huge differences in people’s productivity. Some people’s productivity went way up, some people’s way down. I hope that all universities are taking a hard look at their finances and giving support to those folks who need it.

Grace: [00:33:50] So as we wrap up, I always like to ask the people who come on the podcast if they have advice for people who might be trying to walk in your career path. I know you’ve given a little bit here, but could you maybe say a few words to students, which there may be plenty now who have become interested in virology. What advice would you give to them as they go forward?

Stacy: [00:34:13] So I think from the beginning, I will tell you that there’s no one right way to be a scientist or to be a virologist. You don’t have to do it like I did it. You don’t have to do it like the person next to you. If you’re curious, just keep trying to go one step farther. Along the way, I think it’s really important to find people who will support you and who will be mentors for you. This was really hard for me when I was coming to this system. Certainly I had them, but I don’t think I used them as much as I do now. And I wish I had them my whole career because I think that would have been helpful.

[00:34:47] I would say, just be curious. Go do what is interesting to you. You don’t have to try to predict what will be trendy or popular or important. If you like the science, you’ll find a way to make it important one day or you’ll leave, and so you want to do something that is cool to you and that can motivate you, because I think we all want to be happy in our jobs. And if you don’t love what you do, then you should get another job.

Grace: [00:35:14] Well, thank you so much for your time. Thank you for sharing your thoughts on virology and on scientific culture, if you will. Yeah. Thank you for coming on the podcast.

Stacy: [00:35:23] Thanks for having me, Grace.

Transcript of Episode 37: Jason Arnold

Disclaimer: Transcripts may contain errors.

Grace Ratley: [00:00:00] Welcome to The Bioinformatics CRO Podcast. My name is Grace Ratley, and today I’m joined by Dr. Jason Arnold, who is the technical director of the Microbiome Core at UNC Chapel Hill. Welcome, Jason.

Jason Arnold: [00:00:12] Thanks. It’s great to be here.

Grace Ratley: [00:00:13] So, Jason, tell us a little bit about your research and your role as technical director of the Microbiome Core.

Jason Arnold: [00:00:20] A lot of universities don’t have these sorts of facilities. I know where I did my PhD. We had nothing like this. It was a relatively small I don’t want to call it a liberal arts college. It was science based, but the research infrastructure was limited. So anytime we needed to do any sequencing, we would have to outsource that sequencing to either somebody else in our department who had the facilities to do so, which was very limited, or we would have to send it to a company and have it done there. So at UNC, we have all sorts of different core facilities and these core facilities are accessible to anybody at UNC as well as anybody worldwide really to be able to come to us and say simply, Look, I have these samples. I’m not a microbiologist. I don’t know how to do this, Can you help us? And we provide the expertise in terms of development of their experiment, designing which direction it should go and what they would really need to do, as well as the technology. So we have access to all of these fairly high end machines for automated DNA extractions and library preparations and sequencing and all of this stuff. So we’re able to utilize our infrastructure to help others inside the university and outside the university to reach their research goals and to be able to accomplish what they’re trying to do.

[00:01:33] And it’s not academic exclusive either. I actually have a friend who’s part of a startup company and he often comes to us and we discuss like, Well, what can we work on together? R&D for things that he works with. And this all goes through UNC systems and we’re able to really work collaboratively with industry as well as with academics. And it’s a really great experience because it gives me the exposure to all sorts of different things that I would never, in my own research ever run into. So if my focus is on aging or if my focus is on a specific bacteria, I’m never going to be thinking about veterinary medicine or microbial interactions in primate nests in Africa. Like you don’t think about these things, but these are the projects that come in and we then get the opportunity to work alongside these other researchers and ensure that they get the best results they can get. And hopefully that data gets published. A lot of times it does and it’s affordable as well as opposed to having to buy sequencing machines and all of this stuff, they can just bring it here, use the infrastructure we have in most cases, just the cost of reagents.

Grace Ratley: [00:02:38] It’s very similar to the business model of The Bioinformatics CRO. We’re kind of like a core facility for genomics.

Jason Arnold: [00:02:46] So my personal research, which I do at the same time as I work with a lot of these collaborators, I focus on the impact of the gut microbiome on host physiology in the context of aging. So what I look at is we use animal models for aging and we look at the physiology of the gut and what happens as these animals start to get older. We start to see physiological defects and associated with those physiological defects. We see distinct changes in the composition and function of the microbial community. So what we’re interested in is seeing how we can control that microbial community and modulate the composition of what’s there to be able to reverse some of the physiological defects associated with aging. One of the big ones that we’ve focused on in our research is intestinal permeability. So as you get older, the barrier of the gut starts to fail, starts to fall apart, and you get more small molecules and invasive bacteria being able to pass through into the system. But with specialized prebiotic compounds and modulation of the gut microbiome, we’re able to actually reverse some of that defective barrier, allowing for the hosts to remain healthier for longer.

Grace Ratley: [00:03:55] Yeah. So tell us a little bit about these compounds. What kinds of prebiotics are you using?

Jason Arnold: [00:04:00] So the prebiotic we focus on most is galacto-oligosaccharides. So what these compounds are is they’re basically a galactose attached to a lactose molecule and you have a chain of galactose is attached to that. Now these are compounds that are commonly found in mammalian breast milk, and they’re often thought of as beneficial to infants as the infant is growing. It helps with the early development and colonization, with the initial microbiome these animals have. These compounds interestingly are not digested by the host. So if you, for example, were to eat a bunch of these galacto-oligosaccharides, if you didn’t have a gut microbiome, they’d pass right through you and have very little impact. We don’t know for sure that it’s no impact. We’re actually doing some work now to understand if there is a host prebiotic interaction without the presence of the microbiome. But generally speaking, these compounds have to be broken down by the microbiome or by members of the microbiome in order to have an impact on the host.

[00:05:02] So one of the things we like to look at is the specific genes present within bacteria that are able to break some of these compounds down. So galacto-oligosaccharides are generally metabolized by Galactosidase, β-Galactosidase, α-Galactosidase. And what happens is the bacteria will either internalize the prebiotic compound, break it up into bits that it can use and then secrete what it can’t or the bacteria will secrete these enzymes that are able to interact with compounds in their environment, break those compounds down into components that they have binding receptors for. They can bind to that component that they can use, internalize that and use that. So interestingly, there’s this network within the community where you have a primary bacteria that’s able to take this prebiotic compound, break it down into whatever components come out from that bacteria. Those components, those byproducts are then able to be used by other bacteria in the community. So you form this cross feeding network where, let’s say lactobacillus, for example, will internalize a galactooligosaccharide, break it down into lactate, release that lactate, taking the small simple sugars off that it can use for itself. That lactate now goes out into the community where another bacteria like let’s say Roseburia, can internalize that lactate and potentially convert it to something else, adding residues to it, taking residues off. Ultimately, what we are interested in looking at is how these compounds are broken down and repurposed into short chain fatty acids. Those short chain fatty acids then directly interact with the host providing either benefits or growth factors or well, the ability to grow. So it’s a very complex system. There’s a lot we don’t know yet and there’s a lot we’re working on. So yeah, that’s the small molecules we use.

Grace Ratley: [00:06:48] Yeah, you bring up a lot of really interesting topics to discuss here, but just for our listeners, could you give us a little description of what Prebiotics are and how they compare to things like probiotics or synbiotics?

Jason Arnold: [00:07:00] That’s right. So a probiotic is a live microorganism that when you consume a bunch of them and you have enough. It’s supposed to have a positive impact on the host. Now, not all probiotics do, many do. That’s a whole different conversation for a whole different day. Prebiotics are indigestible carbohydrate compounds, usually fibers or complex sugars of some variety that the host can’t break down. They can’t digest it, but the microorganisms can. And what these compounds generally do is they promote the growth of beneficial microorganisms hence being called prebiotics. Synbiotics are a combination of a probiotic and a prebiotic. So think about it as you’re eating yogurt and your yogurt has lactobacillus and bifidobacterium in it. Beneficial microorganisms with enough of them in there that you’re eating it, it’ll have some potentially beneficial properties. Now if that yogurt was engineered in such a way that it consisted of lactobacillus bifidobacteria and indigestible carbohydrates that those bacteria can utilize to produce additional small molecules that are potentially beneficial to the host, That would be what would be considered a synbiotic because you have a mixture of a prebiotic and a probiotic, and the idea is that there’s synergy, hence symbiotic between these two compounds. Now there’s been a lot of studies done that suggest that the specificity of the prebiotic and the probiotic in order to actually function in a symbiotic setting is extraordinary. It’s very difficult to find a prebiotic probiotic pair that actually does produce an effect that’s greater than just the sum of the parts. So the idea of probiotics and prebiotics is if you’re consuming these things, you’re doing so because you’ve expected it’s going to provide a benefit to you in some way.

[00:08:49] However, it depends so much on what’s already in your microbiome, what bacteria are already there that you may or may not see an effect, and you’re going to have differences between individuals. There’s going to be host genetic factors that play a role. So there’s a lot in terms of the context of what’s really there in the host. Now when a lot of diseases you end up having a dysbiotic microbial community, which means that it’s imbalanced in some way. You have bacteria that are there in higher abundance than they should normally be, or that they would be in a healthy person. And sometimes those imbalances will cause drastic differences in a person’s response to a therapeutic or to a probiotic or to a prebiotic. So one of the things we’re interested in and that we try to do a lot of work with is understanding how how these dysbiotic communities come about. For example, in aging, we know as an animal gets older, these communities start to shift and you start to get this dysbiosis, which makes it harder for certain drugs to impact the host. We’re trying to understand a little bit about how that happens. Is it because diet is changing or is it because all of a sudden people are starting to get older and they’re taking more antibiotics and that tends to skew the community’s dynamics a bit. There’s a huge amount of information out there, but there’s still a lot that needs to be done to be able to really understand what’s going on.

Grace Ratley: [00:10:10] Yeah, the translate ability is really interesting to me. So what topics within microbiota science are you really excited about or do you think we’ll see the most advances in the next five years?

Jason Arnold: [00:10:23] I think in terms of what isn’t currently easily accessible and available is the idea of personalized medicine. We’re moving more toward systems where we’re able to understand the host’s genetic background and how the host’s genetic background interacts with the microbial community. Historically, we haven’t had technology that allows us to do that. Now, with the advances in cell culture techniques and with the advances in next generation sequencing, we’re able to actually start designing models and experimental systems where we can, even from an individual host in the hospital, take a biopsy from intestine, grow stem cells from that biopsy and be able to actually generate host specific organoids, for example, is a system that we’ve done a lot of work in, and those organoids, they’re like micro organs. It’s very hard to describe without showing a picture of them. Basically what happens is you take a subset of primary tissue from the host, which could be a biopsy, it could be a tissue section from the surgery, grow them in specialized growth conditions that allows for those stem cells to separate from everything else that’s there and differentiate to form a what we like to think of as a miniaturized organ. So you’re going to have all the different cell types that are in the small intestine or large intestine, depending on where those stem cells come from. And you’re going to have that host’s genetic background. It’s just ingrained into the tissue because you’re using it from a host.

[00:11:47] Now what we can do is we can actually take complex communities, individual microorganisms, individual small molecules, whatever we want to look at, put those inside this this three dimensional organoid and look directly at host gene expression and host response to that compound or that microbial community or individual microorganism. And that provides us a wonderful tool to be able to start to think about personalized medicine, because these aren’t very expensive to run these sorts of experiments now that the technology exists. So I think that’s one direction the microbiome field is going. Another direction that is actually really exciting to me is historically when we’re thinking about the microbiome and the microbial community and what’s all there in the gut, I’m thinking intestine specifically, you’re sequencing everything that’s there and you’re going to get a combination of bacteria derived from food bacteria that are alive, actively growing in the mucosal layer. You’re going to have bacteria that are just in the lumen and you’re going to have just DNA from bacteria that may or may not even be alive. So you get this big picture of everything that’s there. But in terms of being able to translate that to health, it’s very difficult to do because we have no good way of saying we had no good way of saying, well, these are the bacteria that are actively alive in doing something versus everything that’s there. Now, we have been working on developing and we finally to the point where we’re able to use this technology where we can actually differentiate in sequence data what bacteria are alive versus what bacteria are not alive or DNA. That’s just free DNA.

[00:13:23] So we can take everything that’s dead or excluded from the live bacteria section out and then start to get a picture of what bacteria are actually there. So we ran a study recently in mice, actually, because we were finding that in a lot of the mice that we were using in some of our experiments, we had a huge amount of lactococcus and we did some experiments where we were like, Well, is the Lactococcus there alive? Is the Lactococcus in the food? Like, where is this coming from? And by differentiating live bacteria from non live bacteria, we were able to identify that a lot of this lactococcus signal that was coming up in our sequence data was actually derived from the food. It was not even bacteria. It was just free DNA because portions of the diet were run through a fermenter, a big chemical vat of sorts that produces a large amounts of this food for the animal diet. And there was just lactococcus in there. And we had just lactococcus DNA. It was nothing alive. But in some cases it consisted of like 20 to 30% of the overall microbial community. When we came back with our sequencing data, when we actually looked at what was alive, it was zero. It wasn’t there at all.

Grace Ratley: [00:14:29] I think that would be extremely useful, especially for things like gnotobiotic studies where you’re trying to take a known source of microbes and colonize some sort of mouse with them, seeing how well those microbes colonize within the gut, that could be really useful in eliminating some of the variability that you see in studies, especially with things like fecal microbiota transplants.

Jason Arnold: [00:14:56] Yeah. Also in production of probiotics. This is something that we’re actually thinking about because a lot of times you go to a store and you buy your probiotic and it says there’s ten to the seventh bacteria. Is it alive? So using something like this can help in development of these products. So the people who are actually producing and selling these to other individuals in the hopes that it is going to be some therapeutic benefit, you actually know for sure that these are in fact, alive and you can do tests then. I mean, aside from culturing now, culturing for most probiotics is sufficient, but this is going to be a much higher throughput. When you use this method in conjunction with the PCR, you’ll be able to actually get an exact number of how many bacteria are there in a matter of a couple of hours as opposed to having to wait a couple of days for culture plates to grow. And then you have to take into consideration, well, the plates desiccate too much. And not every colony grew to the point where it has to. Am I miscounting? Is the system I’m using not working right? Being able to do this in a molecular approach, it just eliminates a lot of the the background, so to speak, and a lot of the human error. It will allow for development of much more, I guess, precision in terms of the development of probiotic mixtures.

Grace Ratley: [00:16:08] Well, that’s a really exciting technology. I hadn’t heard of that before. That’s something that you’re working on.

Jason Arnold: [00:16:13] Yeah. So there were some work done years ago that discussed this sort of thing, like basically in environmental samples being able to exclude contaminating DNA. So we repurposed this DNA contaminant exclusion system for sequencing applications. So it’s been done before in, like I said, for like identifying DNA contaminants and for viability of bacteria using PCR based methods. But it had never been done before in a sequencing platform. So yes, we in the Microbiome Core are actually working on getting all of that developed to a point where we have people actually coming in now and using that technology here where it’s actually accessible for people who want to do that kind of research.

Grace Ratley: [00:16:58] So I am very interested in your reptile business. Can you tell us a little bit about that?

Jason Arnold: [00:17:05] Oh, yeah, yeah. So I guess I should preface this by saying I grew up in New York, Western New York on a farm, didn’t have much in terms of technology out that way. I mean, we did, but I wasn’t big into that sort of thing. But I was extremely allergic to most of the animals we had on the farm. So as a kid, I’m like, We have horses and goats and chickens and dogs and everything and I’m allergic to all of these. I’m like, This is horrible. I love the animals. I love being around them, but my eyes are watering all the time. So I must have been like eight years old, nine years old, and I wanted a pet. And somehow ill advised, I convinced my mother to get me an iguana. And I’m like, Now that’s a terrible idea. Never have an iguana as a pet. They’re bad. But little kid Jason was like, Yeah, this is a great idea and it was just out of control from that point on. So I realized, Hey, these are animals that I can interact with and work with and not be allergic. So years later, I after high school, I’m starting my undergrad. I started keeping more reptiles. So I started having some snakes and some other types of lizards. And I started working with other breeders and collectors and zoos and conservationists in the area up in New York.

[00:18:21] And I got to learn a lot about the genetics of these animals and the husbandry and how these animals are taken care of and how exactly you can keep large numbers of them. And as I started graduate school, I started a business where I was breeding a variety of different reptiles for genetic mutations, for colors and patterns, and I was providing them to zoos and to other collectors and breeders. And actually I had produced the first of a lot of different species. My bloodlines are in all different countries, all over the world from years and years ago. So I did this. I also did an educational demo program that I worked with others in the community. So in New York, there were a lot of really strict licensing requirements to be able to have anything venomous or anything potentially dangerous. So I had all those licenses and I had acquired them over the years, having worked with a lot of these people and working in conjunction with high tier rescues and organizations that do a lot of work with this. And I had started a business where I was doing educational demos for schools and for small outreach programs where I would be able to show pretty much every native animal in the United States, like all the different native venomous snakes.

[00:19:34]

So we had like copperheads and water moccasins and all of these different types of venomous snakes and we had crocodilians and we had all of this, you’d think it would like a zoo and we would go from school and we’d show people these things and we’d educate them and teach them best practices in terms of like, if you were to encounter something like this in the wild, how would you interact with it in a way that is safe. And in most cases, it’s just don’t even go near it. Just stay away. It’ll stay away from you. But it was really a great experience. And I learned a lot about the animals themselves, the behavior of the animals. And then after finishing graduate school, I dialed it back a little. I still have a very large collection of a variety of different snakes and gecko species. My breeding projects have all been turned off because I’m focusing on other bigger things. But yeah, I actually still have some eggs incubating right now from some of my geckos. I had a couple of baby geckos hatch last week. It’s a great hobby. I enjoy it as a hobby. The way I thought of it is, I started it as a hobby and then it transitioned into a business.

[00:20:36] And then when I started feeling too much like a business, I was like, Huh, I’m going to dial it back and it’ll be a hobby and I’ll just do this as I enjoy doing it. It overlaps a little bit with my research because there’s a lot of very interesting metabolic potential in reptiles that we don’t often think about. For example, I have snakes in my collection that will go a year at a time without eating. Well they will eat one meal and they will sit for a year. They won’t lose any weight. They won’t behave any differently, but they won’t eat. So this led me to believe like, well the metabolism. Something is absolutely fascinating about their metabolism. And if we can somehow repurpose that microbial community or the function of that microbial community, there’s a lot of potential in terms of what can be done to increase yields in agriculture or even in some cases, help with a variety of different human diseases. If it’s something that we could use to control metabolism, slow metabolism down, if it has to be slowed down, there’s a lot that can be learned from these animals that I think a lot of people don’t take into consideration.

Grace Ratley: [00:21:38] Yeah, certainly. I was just about to ask if you regularly sequenced their microbiomes.

Jason Arnold: [00:21:44] I’ve wanted to. I haven’t gotten funding for that yet. Presumably at some point I will actually do it. I have enough animals in my collection now to generate pretty robust data from it, but what I would probably do is I would network with others and other organizations that I’ve worked with, get samples from a variety of different housing sources and make sure that we have a wide variety to be able to get some decent statistics out of it at least.

Grace Ratley: [00:22:09] That’s so fascinating. So is that how you got into science through reptiles?

Jason Arnold: [00:22:16] Well, it’s a tough question. I’m going to go with No. I think I’m more so got into reptiles because I was interested in science. Some of the earliest things I remember when I was a kid is I would have picture books of animals that live in specific areas. And I would always tell my mom and my family, I want to be a marine biologist. I’m going to be a marine biologist when I grow up. We lived nowhere near the ocean. Western New York, there isn’t an ocean near there. So needless to say, that didn’t happen. The life in the ocean fascinates me, like biology in general has always fascinated me. My undergraduate was actually predominantly in chemical engineering, so I went for studying chemical engineering for three and a half years. I actually had one semester left before I transitioned out into genetics, and it was because I took a genetics course and that was when I was starting to think about reptile breeding. And so I guess it overlaps a little bit. I wouldn’t say that the reptiles got me into science or that necessarily the science got me into reptiles, but they go hand in hand very well. So yeah, after taking one genetics course, I was like, All right, I’m going to put all that money that I could be making as a chemical engineer aside, and I’m going to do science. So that’s how we ended up a biologist.

Grace Ratley: [00:23:31] So then how did you move from genetics to microbiome science?

Jason Arnold: [00:23:36] Then in my undergrad, I was studying genetics and cell biology, and I took a couple of microbiology courses just because they were required. And I thought it would be interesting to understand I mean the smaller side of things. After taking one of those courses, I enrolled in an undergraduate research program, and this is the first time I had done research. And it was this professor who worked on fungal development. So I worked with him for a summer and I realized that A, I really like this, and B, I was pretty good at it. So I was like, okay, this is good. This is a good fit. But I had never even considered graduate school at that point. I was thinking, okay, I’ll finish a bachelor’s degree and I’ll go be a lab tech or do whatever it is I want to do. After doing research for one semester and one summer, I was like, No, no, no, I’m going to have to go to graduate school because this is just what I want to do. So I joined the graduate program, and the program that I was in required us to do rotations in multiple different labs. So I was thinking to myself, Well, I’m just going to stay in this lab doing what I’m doing in fungal biology. This is great. But I was forced to go rotate another lab. So one of the labs I rotated with was a biochemist, like an old school biochemist. It was all the polyacrylamide gels for gel shift assays and all of this stuff from ages past that we don’t really do very often now.

[00:24:54] And that professor was starting a small research project where they were interested in understanding how bacteriophages these viruses that infect bacteria allow for the production of toxins that cause disease in humans. I was very interested in that. And one of the big questions he was asking, because the idea was that he wanted to study the evolution of these toxins. And the thought process was that these toxins were evolving as not so much as a mechanism to make people sick, but as a defense for the bacteria against either environmental factors or environmental predators. So during my rotation project, I had started with this guy and he had told me, All I want you to do is just get a picture, a microscope image of an amoeba eating a bacteria. That’s all I want. And it seemed pretty straightforward and it seemed easy enough. But I kept getting pictures of dead amoeba and I was like, I don’t understand why they keep dying. This doesn’t make any sense. So I met up with my advisor and I was like, Look, I’m using this bacteria that the other, the graduate student gave me and he said, This is the bacteria that he’s working on. It should be fine. The amoeba should eat it, what’s going on? And he was like, Oh, that’s interesting. He gave you the wrong strain. He was supposed to give you a strain that was a knockout for the toxin gene in that strains not but the literature shows that amoeba are immune to this toxin. What’s going on? So we switched the strains immediately, got beautiful pictures of amoeba eating bacteria. No problem whatsoever. Switched back. All the amoeba are dying again.

[00:26:29] So my entire PhD thesis was built around the concept that this paper from the 70s was wrong. And this publication suggested that the toxin that this bacteria was producing, so it was Shiga toxin. It had shown amoeba is immune to it. And they did all of these complex assays where they took pure toxin and they incubated amoeba with it and the amoeba would live no problem at all. And we ended up finding out that, yes, amoeba are immune to the toxin in its pure form. However, if the bacteria is producing the toxin from inside the amoeba, the amoeba will die. So it was very interesting to find that out. And what we ended up doing is we did a lot of molecular research and we did the actual molecular mechanism of what was happening in the amoeba. Unlike mammalian cells, lack the cell surface receptor that’s responsible for binding to the toxin subunit that allows it to get into the cell. So pure toxin is just going to bounce off the amoeba as if nothing is there. However, when the bacteria is already internalized and it’s in a vacuum inside the amoeba and that toxin is released from inside that vacuole, it passes right through the vacuole wall into the cytoplasm and kills the amoeba.

[00:27:36] So we did a whole bunch of experiments on that. And then we actually went a step further and even identified the mechanism by which the amoeba were identifying what bacteria is, food and what isn’t, which was very interesting. And it led to my thesis, which focused on the evolution of bacteria in the context of antipredator response. And that to me was fascinating. So we have this polymicrobial interaction, and I was thinking about that as I was getting close to graduating and I’m like, Yeah, polymicrobial interactions are really interesting. What about very complex communities? What about the gut microbiome? And this is around the time where next gen sequencing was emerging at a point where it was accessible and it was affordable, and we could actually do high throughput sequencing studies and metagenomics was just coming up the conference. I was at my final year of graduate school was, I mean, all metagenomics [] the new thing it was the awesome fun thing that everybody was doing. And all of this data for complex communities was being released. And I said, You know, that’s interesting. I was on the toxin producing pathogen side. Let’s take a walk over to the other side and look at how these beneficial microorganisms interact with the microbial community to help benefit hosts. So that’s how I ended up in that.

Grace Ratley: [00:28:52] That’s an awesome journey and Amoeba is just mind boggling. It makes me wonder a little bit about within the microbiome. Obviously the most commonly studied microorganism is bacteria, but you also have the microbiome, which is the fungus that live in our gut and the virome, the viruses in the gut. Have you done any work with these polymicrobial interactions in your microbiome research?

Jason Arnold: [00:29:20] Yes, I have an active collaboration now. We’re working on finishing our first publication, looking at Protozoal pathogens in the gut in human cohorts, and how these pathogens not only cause disease, because in some cases the disease itself isn’t really noticeable or isn’t really even measurable, but how they interact with the microbiome in such a way that I don’t want to call it symbiotic because it’s the opposite of that for the host. But how these interactions between protozoal pathogen and microbial community lead to more severe disease in the host. So we have some studies ongoing there. And then parallel to that, we have some studies looking at the same protozoal pathogens and how probiotics can be used to combat infection and how specific bacteria colonizing specific niche space within the large and small intestine can displace these pathogens and help clear the infection more rapidly than in regular, conventionally raised animals without probiotic treatment. So we are doing some work there. In addition to that, with technology increasing and the new technologies emerging and becoming more affordable, most microbiome studies historically have been focused on sequencing one specific region of a bacterial gene that’s conserved across all bacteria, the 16S ribosomal gene. So a lot of sequencing studies do is they amplify that region and they sequence it and that’ll give you a snapshot of all the bacteria that are in databases and in that community. So as we advance and as sequencing technologies get better and more affordable, we’re able to get much higher depth.

[00:31:04] And instead of just simply sequencing one gene or one fragment of a gene, we can sequence the whole genome of not only the bacteria, but everything that’s in the community at once. So a lot of the studies that have been coming through recently, within the past 2 or 3 years or so, it’s transitioning more toward this whole genome shotgun sequencing approach where you’re sequencing every bit of DNA that’s there. You’re going to get some host, you’re going to get some bacteria, but you’re also going to get all of the DNA viruses and protozoal pathogens and other any other eukaryote that’s there. Right now, the limitation to a lot of those studies is that the databases just simply aren’t good enough. They don’t have enough information there, especially from the protozoa. That’s one that the databases for protozoa pathogens are so lacking that even from whole genome shotgun data were in many cases unable to even identify the presence of many protozoal pathogens. Viruses are the same thing because viruses are so diverse that there’s not really a well curated database for viral DNA sequences. And unfortunately, we’re not at a point yet where we can well, I mean we can, but not affordably be able to sequence both all of the DNA and all of the RNA and decipher like what RNA viruses are there. So virome studies within the gut microbiome, we’re not there yet. We’re on our way. But hopefully within the next couple of years we’ll be able to actually have technologies that work for that.

Grace Ratley: [00:32:31] What other ways can we improve microbiota science? Like how can we improve the rigor of the studies? Or what needs do you think are unmet in the microbiome science?

Jason Arnold: [00:32:46] There are quite a few. So rigor and reproducibility, at least for the NIH and for most people who work in the microbiome are key factors to keep in mind whenever you’re developing a microbiome study. So one of the big things that is difficult in microbiome work is being able to take your results and replicate those results in somebody else’s lab because tiny things that seem like they won’t make a difference, the impact that those tiny things actually have on the overall output of the data that comes out is overwhelming, diet being a big thing. Like if you have a big human study, you can’t really control for diet in a large human study. So it makes it very difficult to reproduce a human study because the diet of your cohort may not be the same as the diet of the cohort from the original study. In mice, that’s a little easier because you could use defined diets. But things as simple as how the mice are housed can completely change the outcome of a microbiome study. Let’s say you have a study with ten mice and you have the mice housed individually in ten cages, and all ten of those mice are being fed the same diet. And then you run the exact same experiment, but you don’t have the housing available for ten cages. So we’re just going to house them in pairs. You can get completely different results because something so simple as the stress that the animal is under being housed singly versus being housed in a pair or in a group that could result in the production of of stress hormone factors that modulate the microbiome and change what happens in the community. The batch of food that you can use can be different.

[00:34:19] The animal’s genetic background could be different the way they were housed originally. It’s endless. The number of different things that could go wrong, not even necessarily go wrong, but could result in a different result than you were expecting. One of the big things we try to focus on is standardization of our methods. So whenever we start an animal study, we standardize all of the animals we co-housed everything in groups for a couple of weeks prior, all being fed the same diet. Then we split everything. So you have this homogenized microbiome before you start. Now the homogenized microbiome in our study may end up being different than the homogenized microbiome in a different study, but at least starting at a normalized point will give you an idea of what changes are occurring. So those will allow for comparisons on that front. Even though you may see dramatic differences over the course of the entire study, we always try to make sure that the genetic background is clearly recorded and that we know exactly what we’re working with. And when we’re doing an animal study, when it comes to human studies, the only thing you can really do is just increase your sample size, get to a point where your sample size is large enough that any confounding factor that gets introduced is just going to be lost in the background. That’s often very hard to do because a lot of these studies are not cheap and trying to get to a point where you have 12,000, 15,000 humans, it’s very difficult to do.

[00:35:37] So yeah, there’s a lot that needs to be taken into consideration in terms of reproducibility, also randomization of samples when you’re preparing DNA isolation. So you have your samples, they all come in, you’re ready to start your extractions for the sequencing and you’re going to have to do it in three batches. All right, the first 40 samples here, next 40 samples, next 40 samples, run them all. You’re going to have batch effects. So if you don’t randomize all of your samples first, you may see differences between sample groups that are in different extraction subgroups. And those differences could potentially give you false positive results or false negative results. In addition to that, you want to do the same thing again when sequencing. So sequencing batch effects are a real thing. So if you have one sequencing run on an OVA sequence and you have a second sequencing run on a different NovaSeq, even if it’s the exact same samples, the results are going to be slightly different. So what you want to do again is make sure that when you’re loading your samples that you have biological replicates all randomized within all of the different sequencing runs. You have to run library preparation matters. That’s another thing that people often don’t think about.

[00:36:48] The kits that you use or the reagents you use are going to differ from different lots, even from the same company. So if your reactions are being all prepared from a variety of different lots, you’re going to want to make sure that you have all of your biological replicates separated across all the different lots that you have. Otherwise you may introduce biases. And the last thing that we think about when we’re preparing samples for clients or for ourselves is the hands of the person doing the experiment matter. So me making a library versus you making a library, it may be the exact same protocol, exact same reagents, but they’re different. And it’s just a matter of a person’s technical handling of the materials. And those differences all matter when it comes to reproducing a result in a microbiome study. So we try to we think about it in the Core as the fewer cooks in the kitchen, the better. So if you have one person who takes the project and it’s their project and they’re going to run through the whole thing, nobody else is going to be involved because then you’re not introducing all these additional biases as you’re going through the process. Standard operating protocols are always very important as well, making sure that everybody is using the same protocol all the time, but for the big things that we think about.

Grace Ratley: [00:38:06] Yeah. And then microbiota science has become extremely popular within the general public. People are very interested in improving their health using microbiome science. What misconceptions do people have about microbiota science, or what ways can people get reliable information about microbiome science?

Jason Arnold: [00:38:29] So, yeah, this is the age of misinformation, after all it seems in some cases. So it isn’t easy. And because this is emerging as a new popular topic, you have a lot of mixed reviews on what’s real and what’s not and what’s good and what’s bad. And it’s very hard to really know. Now scientific literature helps, but scientific literature is limited. And not everybody has full access to scientific literature. People may be able to get access to abstracts and be able to read it over. Oh, lactobacillus, such and such is beneficial for colitis. Great. But what are the results actually show? Because sometimes people are publishing things and the publication and the title and the abstract don’t necessarily match exactly with what the results are showing. So there is a bit of scientific rigor that has to go into understanding really what’s happening now. I think for the most part, news outlets, they don’t go too deep into the science often. Maybe that’s good, maybe that’s bad. It’s really hard to say. Unfortunately, the information that the general public has easy access to isn’t always complete. So generally speaking, eat a healthy diet. That’s a good thing. There’s no downside, even if it isn’t scientifically proven to change the abundance of lactobacillus and whatever. There isn’t a downside to eating less fast food and eating more healthy fresh food.

[00:39:54] Things like that are good. There’s no real downside to most probiotics. Most probiotics are not going to cure cancer. They’re not going to make your hair grow back and they’re not going to do all these magical things that some of the manufacturers want them to. But adding beneficial microorganisms to your gut won’t hurt. So that’s the way I generally think about it. And I try to think more along the lines of instead of what is the magic bullet, what’s going to make everybody healthy. It’s more along the lines of trial and error for yourself, because personalized medicine isn’t really to the stage where we can say yet. So if it works for you, great. But it doesn’t mean it’s going to work for everyone. So FDA doesn’t work on probiotics and prebiotics at this point. They don’t do any approvals. So any claims that are being made by these companies that produce these mixtures or compounds, it’s unfounded in many cases. I mean, there may be scientific literature that backs it up, but nothing is federally regulated. And that’s something hopefully in the future we move toward thinking about regulations because you can make a claim and have it be based on nothing. That’s kind of problematic.

Grace Ratley: [00:41:02] Do you cook a lot of fermented type foods or drink kombucha type of things?

Jason Arnold: [00:41:07] So you would think I would. I don’t like the taste of it, so I don’t. My fiance does. My diet is predominantly pasta. It’s really unfortunate because you would think with all of this background in microbiome and understanding like what diets you really need to have. I eat a lot of fresh fruits and vegetables, I fresh make my food every day. I don’t eat out very often, but it’s not like eating gallons of yogurt and drinking kombucha or any other random fermented stuff just because I don’t like it. It’s not because I don’t think it’s good for me. I’m sure it is. But it’s just a taste preference and I just can’t get over that for myself, even though I know it’s probably the right thing to do.

Grace Ratley: [00:41:49] It’s totally valid. I’m a fan of Kombucha, but I don’t eat a ton of other fermented type foods.

Jason Arnold: [00:41:56] So on that note, we actually did do fermentation of ginger beers for a while. Like a bunch of us in our group, we thought it would be fun to just make our own ginger beers. So we were actually making our own. We got a starter culture from a friend of ours who was doing this for a long time and we made by modulating the sugar that went in. We fermented different like I made a melon beer and I made just a variety of these different ginger beers and they were delicious. So just like as a hobby. And I wouldn’t like ferment foods full time, but it was all right.

Grace Ratley: [00:42:29] Yeah, I see. I love that aspect of the science. I mean, yeah, there’s a lot of misinformation out there and microbiome friendly hand soaps, which are kind of ridiculous.

Jason Arnold: [00:42:39] Sure, yeah, it is a little silly.

Grace Ratley: [00:42:41] But I do think it’s interesting because it gets people interested in science. It gets them interested in making things or eating healthier. And I do appreciate that about it being popularized. So I guess to wrap up our episode, I normally ask if you had advice to give to people who were trying to go into your career path, people maybe who are at a postdoc stage or early career scientists. What advice would you give to these people?

Jason Arnold: [00:43:13] There’s a lot. I guess the first thing that I would say is probably my biggest regret is, I never learned how to be incredibly proficient at coding. If you have the opportunity to learn how to code in R and Python or whatever, do it, it will benefit you beyond your imagination down the line. I have a very limited experience with it and like I said, that’s one of my biggest regrets is not really going harder into that. It allows you to do a lot more. It unlocks a lot of tools that you may not have access to otherwise. As far as early career scientists like postdocs, graduate students, I guess I would recommend, the biggest thing that I think I learned throughout my PhD was accepting what you don’t know. That’s always something I tell all my students. If you don’t know the answer to something, that’s not a bad thing. It just shows you what you could learn. I remember at my thesis defense, actually, one of my committee members was asking questions that he knew I didn’t know the answer to because he wanted to hear me say, I don’t know. And he basically just asked a question. I was like, I don’t know. But here’s where you can find that information or this is how I would go about finding that information.

[00:44:29] And it forces you to think about not only you don’t want to have that idea that you know everything because nobody does, and being able to acknowledge what you don’t know, it gives you the opportunity to grow as a scientist. And that’s a big thing that I took out of my training and my years of doing this is the more you don’t know, the better in some cases because it gives you more to learn. And I don’t know about everybody out there, but I like to learn. That’s one of the things that I enjoy about what I do. And being able to acknowledge that you don’t know something is the best way to be able to learn what you can learn. Aside from that, mentorship is important. You want to make sure that you work with somebody who’s able to provide you with the feedback you need and the support that you need to make the next steps in the career. It’s not easy. It’s a very competitive in academia especially. I mean it’s competitive in industry as well. But in academia especially, it’s extremely, extremely difficult. So without that kind of guidance and support by somebody who has the experience, it becomes a lot harder.

Grace Ratley: [00:45:29] Well, thank you so much for coming on the podcast, Jason. It was wonderful to hear your perspective about all the interesting microbiome science and also reptile science.

Jason Arnold: [00:45:36] Yeah, absolutely.

Transcript of Episode 36: Bettina Hein

Disclaimer: Transcripts may contain errors.

Grant Belgard: [00:00:00] Welcome to The Bioinformatics CRO Podcast. I’m Grant Belgard and joining me today is Bettina Hein, the founder and CEO of juli Health. Welcome.

Bettina Hein: [00:00:08] Thank you for having me on, Grant.

Grant Belgard: [00:00:10] Yeah, we’re happy to have you. So can you tell us about juli?

Bettina Hein: [00:00:13] Sure. Juli is a next gen disease management platform that helps people manage their chronic conditions. We cover complex chronic conditions, and currently we cover asthma, depression and bipolar disorder. We’ll soon have chronic pain and migraine as well as rheumatoid arthritis. And what we do is we bring data silos together and analyze that data to help people find their triggers and give them recommendations to stabilize their condition. So what we bring together is five data sources. All the data from your smartphone, data from connected devices like your scale or your smartwatch. Third, environmental data like air pollution, humidity, sunlight hours, weather. Forth, your electronic health record data. And fifth, what ties it all together is patient reported data. So how you feel on a certain day, how many episodes of shortness of breath you had, how many times you woke up in the night. Just those kinds of things.

Grant Belgard: [00:01:24] It’s fascinating. How do you deal with and process the electronic health record data, given how heterogeneous it is? Or are you focused on markets with large national health care systems? Or how do you approach that?

Bettina Hein: [00:01:39] No, we’re solely focused on the US market. And what we do is we pull in so-called fire data. Fire is a health care information interoperability standard that is really valuable and coming more into play these days. And what we do is we have a fire server where we can pull this data in a JSON format and then put it into our database. And what our users have to do is they have to allow access to their health systems EHR. So if you are, for example at Massachusetts General Hospital, a patient in their system, there’s a patient gateway, you provide that login and then you can pull that data in.

Grant Belgard: [00:02:31] Do you do any QC checks on that or have you seen differences among sources and how reliable the data may be? Do you find it’s more dependent on providers?

Bettina Hein: [00:02:42] Well, all of the data that we collect has noise in it and certain values that don’t work. We only launched our app into the App Store in March of this year, so we’re still in pretty early stages, but we need to clean up that data. That’s always the case and we don’t need every single parameter. So right now what we’re doing is building up our machine learning models to see which data we need to train those and to get significant results that will help people with complex chronic conditions.

Grant Belgard: [00:03:21] And so when you were describing the set of indications that juli covers, it’s quite diverse. I was wondering if you’ve seen any transferability among those indications or if you’re kind of approaching it as every model is trained exclusively on its own and you’re not really carrying over any design principles and so on.

Bettina Hein: [00:03:41] We’re definitely carrying over data. And so the conditions that we’ve chosen are multifactorial and they’re also influenceable by behavioral change. Those are the two requirements that we have and what we try to help people and what we’ve seen in the feedback from our users and our marketing is that people that have these conditions, they go through cycles where they have flare ups and episodes and they’re very keen on figuring out what those triggers are. That’s their suffering. So this is why we’ve chosen these indications. What we see is really interesting that you can pretty much, even if you don’t know what condition they have, you can often see from their data what it could be. And one of the also really interesting early findings is confirming some of the research that our co-founder, Dr. Joseph Hayes from University College London has done, which is really interesting. We can see that one of the very much influencing factors for depression is air pollution. He wrote a paper on this and we did not expect to see this in our data, but we have a partner that brings us via API, very hyperlocal air quality data and pollen count data. And so when we pull this in, we can see that. And what Joe’s research had shown is that when there’s a spike in air pollution, 48 hours later, there are spikes in people in emergency rooms with suicidal ideation and other depressive episodes. And I’m flabbergasted that this early in our data collection of our company, we are already seeing that.

Grant Belgard: [00:05:41] In terms of identifying triggers, how much similarity do you see in that across patients? Is it just dramatically different patient to patient? But you’ll have these averages that maybe stick out a bit? Or do most patients tend to follow a roughly similar pattern?

Bettina Hein: [00:06:00] Well, it’s still in the early stages to say this definitively. There are individual patterns, But what we do see is that standard medical practice is reflected in our data. But we are definitely working on customizing that more, having the models fit the individual as well as fitting the population. So those are things we’re working on right now. And it’s early days, but it’s really exciting.

Grant Belgard: [00:06:30] Do you have any academic collaborators?

Bettina Hein: [00:06:32] Right now we’re running a randomized controlled trial with University College London to show the efficacy of our solution. It’s an ethics board approved study. The trial is registered. And we’re doing that right now. We kicked it off two weeks ago. We’re recruiting for that. We’re doing asthma and depression. So if any of the listeners want to join in on that, you can go to juli.co and take part in our study. So we hope that within the next few months, we will have that study completed and then showing that. We’re also starting to collaborate with researchers at Harvard who also have an interest in these quantifiable aspects of managing mental health, managing other chronic conditions.

Grant Belgard: [00:07:25] And what is this space look like? Who are your competitors?

Bettina Hein: [00:07:28] Well, the most so competitors are disease management platforms like Omada and Livongo, which is now part of Teladoc, things like that. And Mindstrong is a competitor in the mental health space. So we’re a next gen solution of this. A lot of them have focused on diabetes, diabetes type II and also some cardiovascular things. We have left that out because it’s multifactorial, but the treatment protocol is pretty clear, what you need to do to get better from diabetes type II. It’s a big problem, but it’s not quite the mystery that a lot of people that have these other conditions and autoimmune conditions have.

Grant Belgard: [00:08:20] Cool. And I noticed on your website you mentioned DNA analysis as one of the things that feed in. Can you comment on what juli does there and what the plans are for that?

Bettina Hein: [00:08:29] Well, that’s a little bit of selling ahead of what we have. That is part of what we want to do. It’s not built in yet right now. But what we plan to do there is pull in things that show for example, their DNA analysis that show which kind of medication would work for you. Because with these conditions, they’re oftentimes various types of meds that have to be tried out one after the other. So what we’re promising is that we can help shorten the time to stability and thereby on the ups and downs, the highest and the lowest parts are the ones that have the highest cost for insurers and employers. So we’re trying to lob off the tops and bottoms and keep people within a healthy range. And so if let’s say you have depression, what you typically have to do is you have to try out 4 or 5 different medications over the course of a year or two years, three years. We can shorten that by just six months by using multitude of input. We’ve already gained a lot for a patient.

Grant Belgard: [00:09:54] Do you have any plans to engage health care providers directly or insurance companies? I mean, everything I see on the website seems to be pretty consumer facing.

Bettina Hein: [00:10:07] Right now it’s on purpose to have it consumer facing because we’re building the number of users that we have in order to build our data sets to get feedback and it’s free right now. So we’re doing that to really prove out our thesis around this. And then the idea would be that we get a certain fee per life from an insurer or a fee per employee from an employer, and that we’re not quite sure yet where the physician patient interface will be, how much that can be integrated, because those are very complicated things. But the simple idea is that right before you have a provider visit, you can send your physician essentially an output of this is the longitudinal data, since we’ve seen each other last so that they can see in a way of exception reporting. Okay, what happened? The next step would be that a physician or a part of the care team can set alerts for each individual patient and says if they go out of bounds with their score, their juli score or a certain parameter, then we have to engage with them, call them in or have a telehealth visit because something seems to be going off track.

Grant Belgard: [00:11:34] What is the regulatory landscape look like around this? I know nothing of that.

Bettina Hein: [00:11:39] Right now there’s not much. I mean, you can get this FDA pre-clearance, but it’s not required right now in digital health. I think that will come. In Europe, for example, you have a certification that you have to go register your digital health solution and digital therapeutics so that then you can get billing codes to be able to get reimbursement. But in the US, that currently doesn’t exist. And so what happens in the US is that you look at what part can I slip into and which billing code do I fit into so that I can get reimbursed. I believe that that should be more structured, but it’s currently not the case.

Grant Belgard: [00:12:26] And juli is an all remote company, right?

Bettina Hein: [00:12:29] Yes, it is.

Grant Belgard: [00:12:30] Can you maybe tell us a bit about that? I think you were founded after all the COVID stuff started, but was that the plan from the beginning or did your plans have to change as the world changed?

Bettina Hein: [00:12:42] Our plans completely changed as the world changed. So in our company, two of the co-founders have never met in person. I have never met any of the employees in person and I’m an entrepreneur. I have always been. This is my third company. I used to be a sort of Oracle style, one building company type of person, and I cannot believe that I’m running an all remote company. I’ve totally changed my tune, have become from Saul to Paul in this regard. So yeah, it’s really interesting to do it this way. There are some very clear advantages, but there are also disadvantages to having an all remote company.

Grant Belgard: [00:13:31] Can you comment on that?

Bettina Hein: [00:13:33] Well, so the advantages are is that you have access to talent in the whole world. And with that, you also can have comparatively lower salary costs on average. The United States is relatively high wage country. Three of us, four co-founders are in Switzerland, very high wage company. But between us, we cover 13 time zones. So from Krasnoyarsk in Siberia to San Antonio, Texas, we have people in Sochi, in Valencia, in London, in Lagos, Nigeria, in Boston so it’s crazy. I just find it really cool to have this diversity of people from all different kinds of backgrounds. It’s totally fun. And we teach each other some of our languages and talk about we’re having this holiday right now. You don’t have that. So tell us about it, that kind of thing. So I love that about it. And another advantage is flexibility. I’m a busy person. I’m a mom of two elementary school aged kids. I am a star on a reality TV show in Switzerland. I have board memberships and I’m the CEO of juli, so I have a lot of things on my plate. But because we do this remote, I don’t have to constantly be traveling. I can really slot things in pretty tight. And that is great because having like board meeting here and traveling for that whole day was like spent on that. Now it’s two hours and I have time for all this other stuff.

[00:15:29] So I love that about it. I’m working from home right now. The school is right next to our house. Sometimes the kids come home during the day and say, Oh, I forgot this, I forgot that. I can be here and help them. And that’s really wonderful. That makes me happy as a mom because I’ve been a CEO of tech companies for a while and most of their childhood, their nanny would have been the person to do that. And now I can be that. So that’s great. The disadvantages are that there’s not enough this bonding environment. Company culture is really important for a startup because people need to work so closely together and everybody needs to bring their whole self in to the company to work because we have to wear a lot of hats and we have to figure out who has what talents apart from their strict job description. That’s a little harder when you don’t have the cross pollination of being in the same office. You often don’t know that people have this special skill that you didn’t know about. You have to really make time to build a company culture, whether you’re leading your department or whether I do that as CEO, be very diligent about checking in with your people, having your one on ones and seeing where they are. Because if you don’t see that someone’s down in the dumps and can’t concentrate or whatever, seems sad. If you don’t get a handle on that, you can lose people pretty quickly.

Grant Belgard: [00:17:15] What tools do you use to try to mitigate that icebreakers video or Slack? Or what’s your tech stack for communication and company culture building in a remote company?

Bettina Hein: [00:17:29] We’re not currently using any super specialized tools. We use Slack, we use Zoom, we use notion where we have a wiki about the company and you know, all the different things. But that’s pretty much what we run the company off of. We try to have socials. Last week we had one where everybody had to make their favorite dessert from their country and eat it during the social and then explain what it is, why they like it, why it’s special in their country. So that was thing. We played two truths and one lie because those are interesting anecdotes from people’s lives. So they’re just fun things that you can do like that. We’re starting to do more of that. And I am organizing right now a get together for August. So keep your fingers crossed that despite COVID, we will be able to meet all of us somewhere. It’s right now very hard to say. But I think that if you’re a remote company, you should have regular touch points. I even had that when I didn’t run a remote company. My last company was a software company and advertising technology and we had offices all over. We were headquartered in Boston. We had additional offices in New York, San Francisco, Dallas, Chicago, London. And what I did and this was a cost that I incurred, that we incurred as a company, but it was very, very deliberate, is that we flew in people every quarter for our quarterly kickoff and it was expensive, but that’s where the magic happened. We had fun, things that we did. We had group discussions with people that were in the certain area, cross-functional ones. We talked about our strategy and industry outlook. We brought in guests. It was a real all encompassing thing for two days every quarter and people loved it. They were like, that made them so motivated. And once this whole COVID thing dies down, we’ll definitely invest in that.

Grant Belgard: [00:19:53] Yeah, it seems like remote work has been more difficult through COVID than in ordinary times, even though it was forced on people because a lot of kids were home from school who otherwise wouldn’t have been and so on and so forth. What do you anticipate changing as the world starts to get back to normal? Are there any things that you anticipate will get easier and is there anything you anticipate will get harder? What do you think will change?

Bettina Hein: [00:20:23] Well, what will get easier is that you can have the get togethers. And I’m really hoping for that. What will also get easier is that a lot of the normal life resumes where you have after school activities and camps and things like that running again. And I was lucky my kids go to school and in Switzerland and I’m a couple of months a year I’m in the US. But we only had the kids out of school for about three months last year. And then from the fall on, they’ve had full instruction in school, full time. And I’m extremely grateful for that because when they were home and we were homeschooling, it was for my husband and it was a nightmare. They didn’t have Zoom school and it was just really, really challenging. All the companies that we work with needed so much more support. People were like freaking out about getting PPE loans or other COVID things, and it just all had to become real quick doing emergency rounds of funding etc. It was just a nightmare. What I anticipate will get worse is the business travel is going to ramp up again.

Bettina Hein: [00:21:53] And since I have a B2B company, we need to sell to large insurers and employers. That means going places. And that’s hard. And when I founded this company, I wasn’t quite sure how we would do the monetization. I always hoped for a B2C business model, even though I’ve done B2B companies my whole career, but because I always think, Oh, that’s going to be easier, we can do that from any location. But alas, I think I meant to be a B2B entrepreneur. And so that will ramp up again. What I hope will stay is that people are more cognizant of other things that go on in our lives and more accepting of when kids come in and disturb a meeting. That used to be really hard for me. And I didn’t want people to perceive me as unprofessional or understaffed in my household, delegating the children during work things to somebody else. But now people have just gotten used to it, that men and women have families and they have obligations. And so I think there’s been a lot more reckoning around that. And people understand each other’s lives a lot better.

Grant Belgard: [00:23:20] Yeah, so I’ve been working remotely basically my whole career and I would say that’s definitely been my observation. This was weird a few years ago, but it’s totally the norm now and nobody thinks twice. If you have to put yourself on mute and turn it around and say something to a kid or even give them something to do in the office while you’re on a call.

Bettina Hein: [00:23:45] Yeah, we often hear dogs in the background and stuff that before it was this terrible and they have like, how unprofessional. And now that’s changed. You can see here my home office in my background because right now we’re seeing each other on video, not just an audio. So you have a little bit more context around me than if we just met in a random meeting room.

Grant Belgard: [00:24:16] And so talking about your history with B2B companies and so on, let’s explore that a little bit more. Can you tell us about your career path and what influences you maybe had along the way that ultimately led you to found juli?Because I mean, you’ve been in tech for a long time, but I believe this move into health tech is newer, right?

Bettina Hein: [00:24:33] Yes, it is. So I like to brag that I’ve never had a real job. So I founded my first company right out of graduate school. I was a spinoff of the Swiss Federal Institute of Technology in Zurich, maybe familiar to some of your listeners. We were a spinoff of The Signal Processing Department and we made text to speech software. My second company I founded out of MIT, and we became eventually an advertising technology. Software that helped big brand advertisers optimize their video advertising placements on YouTube and Connected-tv and other platforms. So now I’m doing juli, speech technology that goes into mostly cars and cell phones, ad tech, now health tech. Why? Like, what’s the commonality there? Well, the commonality there is that all of these companies used large amounts of data and applied machine learning or artificial intelligence techniques to optimize the results out of that. My first company, we had a lot of speech recordings and we had to make natural sounding voices out of that. And this was at the beginning of this this century. But we already then used neural networks to do that work with a second company flexibility. We were one of the pioneers in YouTube advertising. We were like in the alpha and beta of that even coming up, and we did contextual targeting. So we sucked in a lot of data from different APIs to show our advertisers how they could win in the auction. Well, actually we did that automatically and the Google or YouTube advertising auction. We helped them do organic optimization because we could see what was really interesting to their target demographic. And we could also very interestingly show them the trends that were happening ahead of them actually creating products around that.

Bettina Hein: [00:26:58] So that was another application of AI and machine learning. And now with juli, we’re doing it again. As I mentioned earlier, how I got into it was from personal experience. When I was doing my second company, I had two kids while being the CEO of a strongly growing company. My first child, our daughter was born prematurely and that was because I had had such a normal pregnancy. That was a huge shock to my system and I could not sleep for the next year and a half. It was really, really terrible. And I went to my GP and he said, Well, you’re breastfeeding, I can’t give you anything. I did that twice and I was just going insane. Because as a CEO of a company, you have to be alert and you have to be performant. So what I started doing was that I wore the sleep monitor, which was actually a headband from a company called Zeo, which unfortunately doesn’t exist anymore. But I could monitor my sleep very exactly and see when I woke up and what phase of sleep I was in. So I hacked myself back to health and figured out what to do, despite my doctor not being able to help me. What that then inspired me is to track other things. As things got on, other difficulties in my life emerged and I was cobbling together different sources of data, my digital scale, my sleep monitoring, a journal, write certain symptoms, my electronic health record. And I was so frustrated as a consumer that I couldn’t bring all of this together. And so I thought, well, there we go. If you don’t find what you need in the market as an entrepreneur, you can create a company that does just that.

Grant Belgard: [00:29:03] It’s really interesting. And I see you did a master’s degree after you’ve been this successful serial tech entrepreneur. You went back to school to study comp sci and AI at Georgia Tech. Can you tell us about the timing of that and what your thinking was behind that? And is that something that you found helpful? Is it something you would recommend to others?

Bettina Hein: [00:29:23] I’m a very curious person and I like to learn things, but I’m also kind of a procrastinator and need a more structured environment to bring myself to really get into new materials. So doing a master’s degree is a great way to do that. Yes, my husband teases me that I’m a degree collector because this is going to be my fourth master’s. But whenever I have a break in the action. In 2018, I handed over my CEO duties at flexibility to a professional CEO. And so I decided, okay, what am I going to do? And I’ve always really wanted to have a degree in computer science because back when I was the college age, girls were not encouraged to take those routes, actively discouraged actually from people in academia. And I always thought that I had made a mistake in not going down that route because I was pretty quickly bored with what I was doing. But I’m not a quitter. So I finished all those courses of study. But I’ve always wanted to go further into computer science and this is allowing me to do that. It’s great fun. It’s obviously challenging to do that next to a company and boards and two children. So I take one class a semester and I mostly do that on the weekend. I’ll do half a day on Saturday, half a day on Sunday, and maybe something during the week. And I know that it’s going to take much longer than it would otherwise take, but that’s okay. I’m not in it to get a promotion at my job.

Grant Belgard: [00:31:20] Do you think this will be your last degree?

Bettina Hein: [00:31:23] No. Have you heard of a management guru called Peter Drucker? There we go. Peter Drucker, thanks for showing me that book from your bookshelf. Peter Drucker had a really interesting way of doing things of continually developing himself. And what he did from an early age on is that he had a focused topic. Every three years he would do something completely differently. And I sort of have taken that. And yes, people make fun of me for it, but I don’t really care. I like learning new subjects, and this is a vehicle for me to do that. Will I always do that, who knows, maybe I won’t do another one. But I will certainly continue to have these vehicles that allow me to get deeper into a certain subject matter.

Grant Belgard: [00:32:25] And as we’re running out of time here, if I can ask you, what’s your most contrarian opinion?

Bettina Hein: [00:32:34] My most contrarian opinion, that depends on which country I’m in. In Switzerland, one of my contrarian opinions is that I am for quotas and boards and representation of women in leadership positions. That is highly contested. I know it also is in the United States, but I believe that the government has an obligation to step in when there is market failure and has an obligation to advance our economies by allowing greater participation of women in the labor force.

Grant Belgard: [00:33:21] And how about with respect to the United States?

Bettina Hein: [00:33:24] I don’t know if this is a controversial opinion, but I am perpetually annoyed, like disgusted by the way that the government is structuring immigration. It’s crazy that if you’re an entrepreneur and you want to hire people, create jobs that you continually have to justify your presence in the United States. And I was on a myriad of different visas during my time when I lived full time in the United States, and it just felt wrong. It felt like there’s no appreciation for immigrant entrepreneurs, and that urgently needs to be changed. The strength of the United States is in its ability to attract the world’s talent, and immigration levels are currently at the lowest levels they have been since the 1940s. That is bad news for the United States. That’s bad news for the demographic development of the United States, for the economy, for entrepreneurship, for the international standing of the United States. And I would counsel the US government, Congress, the President, to take urgent action on that.

Grant Belgard: [00:34:56] Yeah, it’s timely. Just yesterday, Harry Hurst, the founder of Pipe, posted on Twitter about how he’s basically stuck outside the US for the next year because he can’t get an appointment at the State Department at the embassy because they’re backlogged from COVID.

Bettina Hein: [00:35:12] Yep. And some of my employees for example, are stuck in the United States because they can’t go home. And I was in that position as well. I could not go home to my home country because not during COVID, but while certain visa things were going on, we were prohibited from leaving or we probably wouldn’t have been able to re-enter. I remember that one of my fellow students at MIT, he had not been home for the entirety of his four years of undergrad. He had not seen his family because he was afraid he was from Indonesia that he would not be able to come back and jeopardize his education. That’s an abomination. You cannot do that to people.

Grant Belgard: [00:36:05] It’s awful. Yeah. What are you most optimistic about?

Bettina Hein: [00:36:10] I am most optimistic about innovation and health care. I think that lots of people are coming in to health care with fresh ideas. I think it’s speeding up. Innovation is speeding up in health care. And that’s going to help people lead healthier lives.

Grant Belgard: [00:36:29] On that high note, thank you so much for joining us. It is a lot of fun.

Bettina Hein: [00:36:31] Thank you for having me.

Transcript of Episode 35: Bharath Ramsundar

Disclaimer: Transcripts may contain errors.

Grant Belgard: [00:00:00] Welcome to The Bioinformatics CRO Podcast. I’m Grant Belgard and joining me today is Bharath Ramsundar. Bharath is the founder and CEO at Deep Forest Sciences and a company that builds AI for deep tech. He’s also written two books, Deep Learning for the Life Sciences and TensorFlow for Deep Learning. Welcome on to the show.

Bharath Ramsundar: [00:00:17] Thank you for having me on. I’m excited to speak with you today.

Grant Belgard: [00:00:20] Can you tell us about Deep Forest?

Bharath Ramsundar: [00:00:22] Yeah, absolutely. So what we do at Deep Forest is to help deep tech companies, often in biotech, but also sometimes in other industries, build out their AI stacks. What this means in particular really depends on the organization in question. In some cases, it’s more say strategic understanding of what AI can do for them or not do for them. In other cases, it’s much more in depth, actually build out an AI sack. We rely a lot on open source tools. I am the lead developer of the DeepChem Project, which creates high quality, deep learning and AI tools for scientific applications. So we leverage the open powers of deep Camelot to help build out high quality solutions for our customers. And in the long run, I think we’re moving towards actually building out more of a product layer as opposed to being something that’s purely consulting based, but that’s still in the R&D phase. But we think we see the future heading.

Grant Belgard: [00:01:16] Can you tell us more about that? Because of course you hear a lot about architectures needing to be bespoke to the problem at hand and so on.

Bharath Ramsundar: [00:01:25] So on the product end, I can’t say too many details and in honest part because we’re now just at the taking prototypes to our friends and customers to get their feedback phase. But I can say something about how we see custom architectures versus out of the box architectures. So one of the strengths that DeepChem brings is that we have something like 30, maybe 40 architectures now that come out of the box, and each of them has many configurable hyper parameters. In my experience, and I might be quoting Andrew Ng from a earlier talk of his a number of years ago, going from no machine learning to any machine learning is say the first 80%. This is often accomplished by a simple statistical model like a random forest going from something like a random forest to maybe a deep model is that 90% boost. And then finally, I’d say the last 10% is where you go from a out of the box deep learning model to a fully customized deep learning model. With DeepChem given that there is just such a broad variety of models, like I think instead of 90%, maybe we get to the 95% fairly out of box. But yeah you do at the end if you really have a scaled out application, want to develop your own deep learning architectures. But for a lot of the customers we work with, they are sometimes even new to machine learning as a way of doing business. So a lot of the work we do frankly, sits at the 80% and then the 95% stage. It’s only a few customers who are already sophisticated that want to really go to that 100%, which we also do to design custom architectures.

Grant Belgard: [00:02:58] And for what kinds of problems is it often economically justifiable to put the resources into squeezing out that last 5 to 10%?

Bharath Ramsundar: [00:03:09] Cases where I’ve seen the last 5 to 10% be justifiable, it’s typically sticking to biotech. If you have a large assay, that’s pretty well optimized. And this is a foundational technology for your company. This can be pretty worth it. In this case, you’re often not say a startup that may be like a later stage startup, maybe early stage company. This is like a critical technology that will be a pillar of your company for the next 5 to 10 years. Yeah, I think it’s worth it to spend some time optimizing it. It can take considerable expense I think to put a ballpark number in people’s minds. I think market rate is probably a few hundred thousand dollars at least to make a custom deep learning architecture for an application. Something out of the box is of course considerably cheaper and the returns will be about 5%. If you have an optimized scaled out pipeline that could be like a steel, a few hundred thousand dollars, maybe that’s millions of dollars in return. But if that math doesn’t make sense for you and oftentimes for early stage companies, these numbers just don’t make sense. I’d say out of the box is your friend, unless you’re of course a deep learning expert yourself in which you can case you can roll something and then make that part of your core proprietary technology, which a number of companies do as well.

Grant Belgard: [00:04:24] So Deep Forest also has a substack where you write recently mostly about aviation and space exploration. Is that a special interest of yours and where do you see applications of AI in those industries?

Bharath Ramsundar: [00:04:40] The way the deep end of the Forest Substack works is that we do about 5 to 10 week tours of different industries. So we started publishing this year. Our first tour was on semiconductors. We did about ten weeks doing a deep dive into semiconductors. Our most recent ten week tour is in aviation. Typically, we move between various industries. We’ve done climate change. We’ve done energy. We will definitely do biotech in the not too distant future. So I’d say for us it’s more, we specialize in really building that deep market understanding of all these different industries. I think one of the powers that AI and Deep Tech really brings is that to quote carefully, some of the technology I’m working with for a customer in the energy battery space is not that dissimilar from technology that I’m using for a customer in the biotech space. A lot of the ideas carry over some of the deep learning architectures. Again, the out-of-the-box 95% like carries over if the custom stuff, of course it’s a different field entirely. But going with our understanding that most people really want up to that 95%, we think that there’s just considerable cross-disciplinary pollination in the aviation space for example.

[00:05:56] I think that CFD solvers for simulating these fluid dynamics and turbulence are, of course, a major mainstay. If you look a lot of federal grants, if you look at what Lockheed, Boeing pay for its fluid dynamics simulators, there’s been recently a surge in deep learning techniques in fluid dynamics. So Google recently had a paper on their new system. They call it JAX-CFD. So they’re using their new JAX Deep Learning framework to actually write a fluid dynamics solver that is machine learning optimizable. I believe that although these techniques are still very early days, it’s going to have a dramatic impact on the way aircraft is designed in the coming 5 to 10 years. And frankly for every industry we’ve done this in, I think this is the case. If you look at semiconductors, computational lithography is going to have a major impact. Google of course released their paper recently showing how they use reinforcement learning to design the next TPUv4 chip. So just given I think hand industry power of these technologies, we see commonalities in application and we see that not just in theory but in practice with real customers.

Grant Belgard: [00:07:06] Where do you think we are in the hype cycle for ML and do you think it varies by industry?

Bharath Ramsundar: [00:07:12] I think that AI for drug discovery I think is maybe even a little past the peak of the hype cycle. There was a lot of froth in that funding market for a while. It’s settling down a little bit, but still a lot of things getting funded. I’d say now AI for drug discovery is probably one of the more market mature applications. The first startups, I think probably pioneers like Atomwise or others got their start 2015, 2014, something like that. So it’s been around for a while. I think a lot of investors now have an understanding for what these companies can do and cannot do, which is that in some cases, if you look at companies like recursion, there actually have been very powerful exits for investors and I think real technology invented. But cancer isn’t cured yet, that to say the least. And as I’m sure that you all would have seen at the Bioinformatics CRO like drug discovery is a hard, hard problem. And we don’t expect AI to really ride in on a white horse and cure anything anytime soon. But I think we’re nearing understanding that these techniques are quite useful in practice, in moderation by a team that knows what they’re doing, which is again to quote that Gartner Hype cycle, maybe we are moving down towards the trough of disillusionment. But then over to the the steady state of useful application. Other industries I think we are much earlier. So things like fluid dynamics, things like high-performance computing, I think it’s just at the early days where people are beginning to realize, oh wait, these techniques are actually applicable to our work.

[00:08:39] So I think that it’s probably several years behind on that funding cycle and I anticipate more hype coming about. There’s also just been I think considerable advances in the technology of that field that are very recent. So I believe there is a recent paper, I think from a Niemann and Kumar’s group at NVIDIA on neural pediatrics, where they take these partial differential equation solvers and use essentially deep learning. The technical explanation, they work in the Fourier space and they make transformations there. But these could potentially speed up solutions of certain classes of differential equations which will just have broad applications in fluids, in energy, a variety of different use cases. So I think that’s probably a foundational paper that will only begin to see play out over the next five years. So yes, to long winded answer very much depends on the field. I think in drug discovery, maybe we’re in for a bit of disillusionment as people realize that techniques are very useful, but they’re not going to cure anything. Whereas in other fields I think there will be just, Oh wow, what if we could design a flying aircraft that’s hyper efficient? I probably hope I don’t say spoiler for everyone, but I don’t anticipate that type of revolutionary advance off the bat in any field. But I think that it is nearing a place of broad applicability where techniques are just useful to people in many industries.

Grant Belgard: [00:10:02] And how do you think about how companies can build that out? I think it’s fair to say there’s a pretty severe shortage of people with substantial experience in deep learning. I mean, obviously there are a lot more people who have some shallow understanding through MOOCs and things like this. But how do you think the landscape looks in terms of employment and training in that space?

Bharath Ramsundar: [00:10:27] I think that the educational tools have really gotten much better. It’s much easier for engineers to just pick up some basic machine learning just looking locally. And my family definitely had a few older engineers who got bored and picked up a Coursera course or two and now do some basic machine learning. These are say veterans of like IT industry later in their careers who are bored with their day jobs. So I think that there is a very positive effect where you will begin to see experienced people in all sorts of disciplines, just pick up some basic machine learning and realize that, hey, this isn’t that exotic. I will also say at the same time though, that in my experience there’s a steep curve not at the early theory stages, but in figuring out how to apply these things in practice. This means that your organization will need to do things like work out infrastructure for where do I do my compute? How do I run this on AWS? Where do I store the data? How do I version control models? How do I keep up to date with all the latest and greatest on the deep learning library infrastructure? While this is certainly possible I think for a medium sized company, it’s very hard I think to keep up with the speed at which the industry moves. With a package like DPM, we are fairly sizable open source package at this point, pretty active developer community. And still we struggle just like the fire hose at Google or Facebook or whatever puts out where they can just keep putting new the world’s best PhDs onto a problem. It just means that it’s very hard to stay abreast of the latest stuff in the fields.

Bharath Ramsundar: [00:12:01] So I think that there is considerable room for good software solutions to play a middle ground. I think teams will want to control their data absolutely. They will want to control their models. They want to not have to have someone who’s the middle man constantly holding their hand in the long run. I think to get off the ground though, having someone is invaluable. But I think what we see as the future is giving teams the ability to run more complex systems, but do so in a way that is as easy as it can be for them. So there’s definitely some solutions from the big players like AWS SageMaker, I think is the one that’s quoted a lot. Unfortunately in our experience, SageMaker is not yet ready really for custom applications. While it does say things out of the box the random forest equivalent, if you look at say just their logging capabilities or how you actually monitor a system on SageMaker, it’s not really at the place where I can recommend it for anyone to use. So there’s a new crop of companies that are like weights and biases, for example, that are offering new developer tools that are beginning to pick up some of this gap. So we think that that’s where we see the future of the broader AI developer market []. For deep tech, we think that there’s a lot of things about scientific projects that are just distinct enough from everyday data science applications that we think that there’s room for product in that space. And that’s the niche that we’re exploring right now.

Grant Belgard: [00:13:33] That’s very interesting. So are there specific areas of deep tech that you think are lagging far behind in terms of application of deep learning?

Bharath Ramsundar: [00:13:45] I would say that the answer is perhaps the opposite in that. I think drug discovery has been a standout for how fast it’s moved in adopting deep learning technologies. I would say that nearly every other field is far behind. The closest I’ve seen in second place is that material science has recently started to see a boom in more machine learning applications. There are some really cool projects, Matt Miner, The Pi-Match and community that have been leading this charge. But if you look at a lot of material science papers, I think that it’s just getting off the ground. Facebook recently launched the Open Catalyst Program to do machine learning on catalyst design. So I think you’ll begin to see a lot more members of that community really uptake these tools as it becomes clear that these techniques work. So that’s probably what I would say is the second runner field right now materials. But yeah if you move past that, it’s very early research right now. There’s a lot of interest. I think a lot of companies are interested in applying these tools. I know that if you look at companies that are doing things like designing cars, again very interested. I think people are very interested in using these reinforcement learning techniques that Google puts to use. But it’s challenging. If you look at reinforcement learning, it’s notoriously finicky. I think Rey has done a great job of making this easier for people, but I’d say it still requires an expert team that really understands what you’re doing.

Bharath Ramsundar: [00:15:11] So there’s this gap where there are things Google can do or OpenAI and then there’s things that everyone else can do, which even for a very solid academic team or frankly a company like ours, we can do things. But we can’t run them on, say 100 TPUs or whatever that Google can toss together. The other players, I’d say the Chinese ecosystem has been putting in piles of money. So Tencent, for example, I’d say is always say two steps behind Google, which is frankly probably five steps ahead of everyone else. So they are innovating in their own right also being fast followers. I will say for non-Chinese companies, there are major downsides so depending on the Chinese ecosystem. You can see the controversy with TikTok or Zoom, I would say, about data privacy and security. But they are frankly innovating and doing excellent work as well. For the rest of us I think we need to figure out how do we maintain cloud stacks, how do we actually scale out our learning infrastructure. And I don’t think there’s a turnkey solution for a new company even say one, I’m going to name a name Lockheed, for example, to waltz in and say I want a deep learning stack. I actually think that will take considerable investment in time. And I’m sure they’ve been doing this already for several years or trying.

Grant Belgard: [00:16:27] So speaking of China and Lockheed in the same segment, is it clear where the focus of applications are in China? Is this something that the military is playing a role in?

Bharath Ramsundar: [00:16:43] There is considerable aggression out of China right now. It’s like the 100th anniversary of the founding of the Communist Party I believe. China of course has comparatively done very well in the coronavirus pandemic, which has boosted its international profile. I would say there is considerable anxiety amongst the military about the capabilities of China. I’m not an expert at AI policy, but I would as a bystander say that in terms of AI advancement, I think the US is doing just fine. Like I see very good work coming out of the Chinese institutions. But I see better work coming out of Google or Facebook or DeepMind, where I think the Chinese ecosystem is ahead is that if you look at their physical manufacturing capabilities, I just read a report this morning about the US Navy is overbooked and a little bit under budget, whereas the PRC Navy has been dramatically accelerating its shipbuilding, which we have a deep into the forest piece about explaining CSSC is the China State Shipbuilding Corporation is the world’s biggest ship maker. And they use the same docks to make aircraft carriers as they do commercial oil tankers. So were I a military planner? And I know there are military planners worrying about this. I would worry about the Navy. I’d worry about the physical hardware. I think in terms of intelligence, I think the US is fine. Like AI, Google and others are doing more than fine right now.

Grant Belgard: [00:18:10] And where is Europe in all this?

Bharath Ramsundar: [00:18:12] Well, that’s an excellent question. I think that Europe has been putting a lot of money into getting their AI ecosystem off the ground. I see a lot of DeepMind of course, but there’s also I think increasingly a number of sophisticated European AI companies. And there are some neat companies doing things in the hardware space. ASML of course has continued to innovate and do excellent work. So I’d say Europe has a fine ecosystem, but in many ways the European ecosystem is you could say a mirror of the American ecosystem in some ways, but there’s a bigger focus on consumer privacy and data protection, perhaps a tad more regulation which is I’d say better for consumers, but maybe a little harder for businesses. So I don’t think Europe is doing badly at all. But I think the rising juggernaut of course is China and maybe a couple steps behind the Indian ecosystem, which has had a lot of innovations I think at the app layer. But not say quite to the same degree as the Chinese ecosystem right now.

Grant Belgard: [00:19:13] Are there any other regions doing notable work in this space?

Bharath Ramsundar: [00:19:17] I think that there’s a lot of interest in Africa. I think there’s been some great deep learning conferences. I think there’s a lot of really talented students who are starting to build out community there. I think the Nigerian tech ecosystem is also booming, for example. So there are definitely innovators in all these spaces. I know less about the South American ecosystem, so I won’t comment there. I know there’s a couple of cool companies out of Brazil. Yeah. So I think that’s probably the other player l like Australia has of course been doing lots of cool stuff in different areas. I think with Australia a big focus right now is there is this ongoing trade war with China where there’s a very challenging situation that they’re facing. So I think the geopolitical side in the Pacific is unfortunately bifurcating everyone where you’re with the US or you’re with China. I think the unfortunate way that things are shaped down there is probably going to be a lot of competition in that entire part of the world in the next coming decades. You asked about our newsletter so I think part of the reason we do these analyses is that I think geopolitics and AI and deep tech are just intricately tied. For example, companies working on better ships, I think will have a very bright future 5 to 10 years ahead when the Navy realizes as it’s realizing right now that oh crap, we have a problem on our hands. There are many people who have been doing these analyses, but what we try to do is look at the broad picture across industries and tie together these thoughts of what we see in different spaces, where there’s actually an underlying theme.

Grant Belgard: [00:20:46] What impact has COVID 19 had on uptake of AI, if any? I mean, do you think it’s had an impact at all?

Bharath Ramsundar: [00:20:56] I will say that it’s broadened the geographic scope to some degree with DeepChem for example, we were a project that grew out of Stanford, my PhD thesis work. The early community was people who showed up at events we put on for the local Stanford community. We put out a pizza. Typically, a friendly company would rent out their space. We’d have a couple of talks, people would mingle. So understandably, our early contributors came from Palo Alto or thereabouts. But increasingly now I think the community is very, very global. So we had some calls this morning, people from Switzerland, people from India, people from Japan. California, of course, remains pulled out. But I think there’s considerable geographic widening where more and more, I would say the AI community is distributed. All the work happens on the cloud anyways. It doesn’t really matter where you are that much. So I think that COVID has accelerated a trend that was already happening by removing the physical necessity of being in one spot. Yeah, this is something that will likely be around to stay. At the same time, I think if you’re an entrepreneur, I think there’s an advantage to just being hanging out in San Francisco, which even today is considerable. So I think both these things are simultaneously true. So I anticipate you’ll probably have a whole bunch of companies where the founders come, set up shop and have staff with the engineers or team or wherever in the world. So then you get the best of both worlds. You get talent globally, but founders locally. And that’s a choice we’ve made. So I’m kind of based in the Bay Area, but the Deep Forest Sciences Team is many places.

Grant Belgard: [00:22:33] Can you tell us more about DeepChem and how it came about?

Bharath Ramsundar: [00:22:38] Yeah, absolutely. So a number of years ago I had the good fortune to intern at Google at their Accelerated Sciences team. So we did some cool work. We trained some deep models that were for the time, I think very cool and very large. I had an excellent internship there. But as with all good things, the internship ended, had to head back to grad school and I realized, Oh crap, my best results are at Google and I can’t replicate any of this. So I set about trying to replicate them. And at the time Francois Chollet had just put out Keras, which was just an amazing tool. So the original version of DeepChem was adaptation of Keras to training multitask networks, which is what we’d built on chemical data. And I wanted to share with some friends down the hall, so put it up on GitHub and made it an open repo and things just grew since. I think the code base has been written considerably multiple times over the last few years. I think our first use case that drew in a lot of people was we had some good implementations of graph convolutions contributed by some engineers who got involved with the project early on.

Bharath Ramsundar: [00:23:42] So that drew in a lot of people into the community. But increasingly today I think that for DeepChem, we are evolving into a AI for science framework. So we continue to have I’d say the best open source suite of machine learning for chemistry tools, I would argue. But increasingly we have very powerful capabilities in materials science, protein design, early work in other fields, been tinkering with some hopes for getting some fluid support off the ground. So where we see the future of DeepChem of going is to really make it easy to apply AI for scientific applications. Deep Forest Sciences of course we support this extensively because, I’m the same person and a lot of the same core thesis drive this. And I think the model we follow is that open core is a just powerful base for any company because you can have technology that’s vetted by scientists and experts across the world and you just get corner cases filled, you get bug reports figured out, you get people contributing their time because it’s open and they can also benefit from it that you don’t get otherwise.

Grant Belgard: [00:24:47] So you also co-founded Computable, right?

Bharath Ramsundar: [00:24:49] Yeah, absolutely. Computable core technology was building out what I would call these data co-ops. The idea behind a data co-op is that if you have a group of people who are gathering a data set, they deserve a right to having some equity in that data set. The motivating example we started from was these genomic patient datasets. If you are a rare disease say patient group and you contribute your genomic data in the quest for a cure, at the least you should be getting some royalties from that. Like maybe not even for yourselves, but so you can continue funding research into your rare disease. So our technology that we built out was to build a system to track ownership of a data set and to parcel out royalties to the owners of a dataset whenever it was used. So we built a system for this on Ethereum before of course this most recent giant crypto boom. But I think it was just a case of cool technology but wrong timing. And we found that the blockchain was just too onerous to use. We had major UI issues. Customers liked the concept, but when they figured out that they had to click seven times to do anything because of complicated back and forth permission granting to Ethereum and it took several minutes for each transaction to go through, it just didn’t get off the ground. So a setback from that team a couple of years back, the team since rebranded, trying a few other experiments. But I think it was a really cool idea we had. Just I think that would be a great project to try again, but say five years from now, once these technologies have matured.

Grant Belgard: [00:26:21] So can you tell us in a nutshell about your path? You grew up in the Bay Area and let you take it from there.

Bharath Ramsundar: [00:26:30] So grew up in the Bay Area. Your first job out of college worked as a engineer at a company called Fusion-io. We used to make these non-volatile flash devices that we sell to Facebook and Apple for quite a markup. This was a proprietary software stack that was very efficient. The company was later bought out by SanDisk. Unfortunately, the commoditization of that hardware market just removed the margins that made that company really, really work out. So I spent about a year there, then left to go to grad school at Stanford. At the time, deep learning was very hot. Got into that, did a lot of coursework, had the good fortune to work with some collaborators who knew more about the chemistry drug discovery side than I did. So learned some of that, did this project with Google and then from there started the DeepChem Open source project. After the PhD, I was more entrepreneurially minded than academically, so I decided to try co-founding startup. At the time, this was in 2017, there was a big crypto boom so we got caught up in the craze and tried building cool things with crypto.

Bharath Ramsundar: [00:27:36] But as I just mentioned that the technology was not ready for I think the applications that we wanted to do. So after stepping back from that company, I decided to take some time off. So I consulted a little bit with a few friends. But the patterns that I’ve been talking about with Deep Forest Sciences became apparent and near the end of last year decided to actually say, okay, there’s something here. Let’s make this an actual company. So that’s what I’m working on full time now. And you’re working to grow out Deep Forest Sciences. And of course, this entire time in the community has been steadily growing and we’ve been maturing the code base and expanding that. At this time, I think it’s something like 70,000 lines of code, and that’s getting to I think a pretty sophisticated numerical scientific infrastructure. We have a long ways to go before really science is so vast. There are so many places software can make a difference I think. So it’s probably another 5, 6, 10, 15 years for this tool to really mature.

Grant Belgard: [00:28:34] Do you think at some point DeepChem may need to be renamed?

Bharath Ramsundar: [00:28:39] We’ve definitely had some discussion about that. Yeah, probably. But I figure let’s build the infrastructure and the communities first and at some point the name will work itself out. But yeah, it’s entirely possible that we need a better name. But for now, I think everyone in the project understands it’s broader, just chemistry. It is something we have to tell newcomers as they come in that, Hey, I know we’re named for chemistry, but we do do that, but we also do other things. So make sure we don’t miss out on that eventually.

Grant Belgard: [00:29:08] What do you expect will be the most visible, successful applications of AI over the next decade?

Bharath Ramsundar: [00:29:17] OpenAI I think is likely to use its GPT-3 technology for some high profile applications. So I think they put out this very recently copilot or something like that where they’re using these GPT-3 technologies to help aid code autocomplete essentially. I think this is going to be a very radically powerful technology. We’ve seen the way software is developed now is very different from the way it used to be. So we all have our continuous integration systems. We have all these automated controls on a software people. It’s what enables a small team DeepChems to maintain a probably 10 or 15 years ago to require a large company to maintain in terms of code sophistication. I think these trends will only accelerate. It’s going to be a case where one, it does become easier for everyone to develop software using AI techniques. But I also think it’s going to be the case that the people who have the best understanding and control of these methods will get richer style, make the most use of it. So I think that this is going to have a dramatic impact on the developer market right now. One thing I see is that if you look at say something like front end developers, if you go back 20 years, web developers were very limited market supply. But as you know, the growth in this technologies has dramatically widened as you’ve had code bootcamps and even increasingly now, I think autocomplete style tools is the next generation of this.

[00:30:49] You’ll start to see a bit of the pricing on that market fall down. So my guess is that what will happen is that you’ll be able to have say one senior developer who can corral AI tools to do more, which might undercut the market for some of these bootcamps over time. I think that something similar is likely to happen in basic data science as well. There has been a lot of basic data science bootcamps. A lot of kids are learning these skills in college. My 2 cents is learning math and stats is something that you can’t go too wrong with and that it’s just such a powerful way of thinking about the world. So I’m not too worried about college students who picked up some extra stats not making use of those skills. But I do think that for everyone in the tech ecosystem, I think we’ll have to continually revamp our toolchains and our understanding of these technologies in order to stay ahead of probably the increasingly powerful AI co-pilots that are coming our way. One kind of example a little bit, this is a field in pure mathematics so Peter Schultz, who’s one of the world’s most foremost mathematicians, recently put out this blog post where there is a conjecture that he’d strongly suspected was true. But had not been able to entirely verify to his liking. But he was able to use Lean, which is a new proof assistant/dependent type programming language from Microsoft Research to formulate a proof that it actually was correct. And he was very surprised that these tools were at the point where it actually could aid in the work of cutting edge mathematics and not just for something you do after to reprove old theorems.

Bharath Ramsundar: [00:32:26] So I think that these trends will only accelerate. I think software will cannibalize software far before it succeeds in cannibalizing. Plumbers for example, I think can probably sit safe. I don’t anticipate plumbing robots coming about for the next several decades. So weirdly, I think you will see a case where plumbers mechanics I think will sit tight knowing that their jobs are likely to be in high demand. Whereas like I think for developers, I think there will be a bit of a struggle where you need to upskill yourself in order to be competitive in this market. And that means probably just learning more mathematics and learning more system software at the high end system scale and at the sophisticated mathematics. I don’t think those skills will go away. But if you only know how to do basic HTML, you might be in for a tough time. So that’s maybe the biggest shift I see in that AI is going to move on the software industry and that’s where I think will have its biggest impacts, which will be I think insulated from the world more broadly. But for those of us who work in the industry, I think we better crack open those textbooks and start learning some more things. Otherwise we’ll be replaced.

Grant Belgard: [00:33:29] What do you think is the best way to go about that for people already in the workforce?

Bharath Ramsundar: [00:33:33] That is an excellent question. I think that MOOCs are really amazing. I think just starting from a MOOC, starting from YouTube, Wikipedia even can be surprisingly useful and doing lots of side projects. As a developer, I think that you have 9 to 5, you have things that you have to do for your work. Like for me for example, the 9 to 5 is of course like this AI deep tech, but I have persistent interests in compilers. That’s something that I toyed with on the side and nothing to really show for it. But I think that these experiments I find often come in and really change the way I do my day job for the better. Having a curiosity and willingness to do these toy projects I didn’t know any react. So I spent some time hacking together, a very simple React app. It was terrible. A good frontend developer could do much better, but I think it broadened my understanding. So I think for all of us, just given how vast computer science is making something that you don’t know much about, maybe like a simple graphics rendering system and doesn’t have to be professional grade. It just has to be a weekend or two of hacking. I actually think that’s an excellent way of keeping our skills. I think our greatest advantage is our flexibility or if you needed to get a good Java developer to code and C++, you could probably figure that out given a couple of months, whereas I think that’s going to be harder for like an AI system to do necessarily. So I think flexibility is where we have a persistent advantage over the machines.

Grant Belgard: [00:35:01] Well, I wonder what do you think of an alternative hypothesis that basically these tools will be developed to aid and supercharge developers. And maybe it’s not so much a matter of having developers put out of work as just having dramatically more software development around the world. I mean, how likely do you think that outcome is?

Bharath Ramsundar: [00:35:23] That’s a really, really good question. I don’t know for sure at all. I think this is such a complicated question that all of these could come about. I suspect that if you look at say web development as an example. So I think that in one way, the number of web developers has skyrocketed. People use Squarespace or similar tools all over the world to make new websites. And you could argue you are a web developer if you’ve built a website on Squarespace. At the same time, the skilled web development market I think has gone in two directions. You have the very high end folks who are building these ultra sophisticated stacks. Yeah, of course they’re around. They are probably getting paid multiples of what they were getting paid before. But I think the city basic website together market has vanished. And it’s so easy with no code or do it yourself tools now that as a new entrepreneur together Deep Forest Sciences basic website on Squarespace and we have a new website that we’re actually building as a custom app. But the Squarespace is good enough for us for getting something up. Similarly, I think for a lot of development applications, we’ll see a similar bifurcation.

Bharath Ramsundar: [00:36:35] I think the current market will be automated out. You’ll have the up skills market where, yeah, it’s going to be very, very lucrative. In one sense, the number of AI developers I think is going to just dramatically increase. But if in the future you say, Hey Siri, can you put together a basic website for me and Siri does it, are you an AI developer? Yeah, in some sense you are. You’ve figured out how to interface with an AI and make it do useful things for you. But that market of the person who would do that right now, either they’ve turned into the Siri developer themselves so to speak, and they’ve upskilled themselves in another option. They’ve decided let’s just use my skills to build out a business, which I think a lot of technical folks increasingly do. I can see both these, but again I should put an asterisk on on these claims in that. I only have one limited view into this very, very broad ecosystem. So I am as ignorant as anyone in terms of who knows what all the dynamics that are happening. It’s such a complicated industry. There are so many things going on, so we’ll have to wait and see.

Grant Belgard: [00:37:37] I think it’ll be a really interesting decade.

Bharath Ramsundar: [00:37:39] Yeah, computer science, I think it’s one of these fields just reinvents itself like all the time. So I think that this is probably a very natural part of the process that’s been going on since the 50s.

Grant Belgard: [00:37:48] Thank you so much for joining us. It has been fun.

Bharath Ramsundar: [00:37:48] Thank you for having me on.

Transcript of Episode 34: Saroja Voruganti

Disclaimer: Transcripts may contain errors.

Grace Ratley: [00:00:00] Welcome to The Bioinformatics CRO Podcast. My name is Grace Ratley, and today I’m joined by Dr. Saroja Voruganti. Saroja is Associate Professor of Nutrition at the University of North Carolina at Chapel Hill and her lab studies Nutrigenomics and Nutrigenetics. Welcome, Saroja.

Saroja Voruganti: [00:00:17] Thank you, Grace. I’m happy to be here.

Grace Ratley: [00:00:19] We’re happy to have you. So can you give us a little bit of a description about what your lab does?

Saroja Voruganti: [00:00:25] Yeah. As you mentioned, we do study Nutrigenomics and Nutrigenetics. We are trying to understand how genes affect, how our nutrients are metabolized, how they are broken up in the body, and how genes affect that process, which is termed as nutrigenetics. The other side we have how nutrients affect how the genes are expressed, which is the nutrigenomics part. So we are trying to understand how these nutrients and genes interact in relation to subsequent disease risk. So my labs interest mainly is purine metabolic pathway. So we are trying to understand how the enzymes work or how genes affect those enzymes, how the nutrients affect this pathway and how it subsequently affects various diseases, mainly studying neurodegenerative diseases as well as obesity, how it affects those.

Grace Ratley: [00:01:23] Awesome. So can you give us a little bit of an idea of how what we eat changes purine metabolism?

Saroja Voruganti: [00:01:29] So a key example I can give you is fructose, in normal levels is fine. The problem comes up when you have too much and then not many people know, but it releases a molecule which can further convert into uric acid, which is a purine and uric acid again, in normal levels is an antioxidant. It’s good for your body, which is an end product of purine metabolism. But I’m talking about all this when it is too much like you don’t realize you drink, you eat fructose containing drinks and food and we don’t realize how much we are taking it in. And it causes more production of uric acid, which if it is too much in the body, it can turn into a pro-oxidant and cause harmful effects. So fructose is only one of them though. But you have other products like any purines, alcohol. Everything in moderation is okay. When I talk about these foods, I’m talking about when it is too much. So that’s why we always insist on moderation and balance in your foods, right.

Grace Ratley: [00:02:36] Yeah. And notoriously high fructose corn syrup is in high concentration in sodas and such. And for our listeners, purines are things like adenine and guanine. So the nucleotides that build our DNA and our RNA.

Saroja Voruganti: [00:02:50] Exactly. Yeah. And most of the organ meats and all contain hyperons. But if you see purines themselves only contribute 10% to the total purine content. The other products like fructose and all also contribute quite a bit.

Grace Ratley: [00:03:05] Why did you get into Nutrigenomics?

Saroja Voruganti: [00:03:08] I’ve always done nutrition. Back in India, I had done my bachelor’s in foods and nutrition, which is called, and then I did a dietetics course there. Then I moved to US for my husband’s job and after several years of gap, I started thinking of going back to school and the only thing I knew was nutrition. So we were in Austin at that time. I went to UT Austin. I thought, I’ll do Masters, but eventually it became a PhD, So I did my nutritional sciences. I’m in PhD in nutritional sciences. We did a study there where we were trying to give an ideal diet to some women and I saw that they were all similar age groups, similar weight. All were women. So sex wise, they were all same.

Still there was so much difference in how they responded. And I started getting interested into genetics, like how genetics plays a role. So I started looking at various ethnic groups as well as within ethnic groups. Also how genetic variation can affect, how people respond to diet. One of my committee members was Dr. [Comuzzi], who is an expert in genetics of obesity. When I finished my PhD, I went to him and said, I would like to do a postdoctoral fellowship with you and understand more about genetics. And he was nice enough. Without any genetics background, he agreed to let me do his postdoc fellowship with him. So that was when I started getting interested. Like how are genes affect, how nutrients are metabolized or how are nutrients affect the way genes are expressed.

Grace Ratley: [00:04:46] Yeah, I find that very exciting. I know that within nutrition research, it’s very common for people to say something is good for you or bad for you, but there’s a lot of variation within the population.

Saroja Voruganti: [00:04:59] Yeah. Based on a person’s gender and race ethnicity. And there are so many other factors. We haven’t even touched the tip of the iceberg.

Grace Ratley: [00:05:10] Yeah, it’s a very complicated field. So you study nutritional science, but your initial degree was in dietetics. Can you give us a little bit of a difference between that?

Saroja Voruganti: [00:05:21] See, that’s what happened right. When I was doing my dietetics, I did study the biology behind it, how the metabolism happens and how you give a diet plan to people. But there’s so many complexities involved in it, like the genetics or now we know a lot about microbiome and there are so many other things involved. Unless you have a very deep understanding, I feel like it’s difficult to convince people to change their diet or even one nutrient. In dietetics, we do understand metabolism and all, but I didn’t feel that it was comprehensive. I think Dietetics plus nutritional Sciences is a very good combination, but I don’t know how many people can do both.

Grace Ratley: [00:06:03] What do you think is needed for us to move nutritional science into applications and into something that people can actually act upon in their day to day lives?

Saroja Voruganti: [00:06:15] I think as we are closing in to this precision nutrition, can we precisely say who needs how much food? We’re working towards it, but we are not there yet, of course, because we need to understand all the genetic variation. Also, we have been only studying single nucleotide polymorphisms in detail, but there is so much other genetic variation, structural variation which we haven’t even touched in detail, like copy number variation and many others. And then we haven’t even looked at gene expression. So that’s what our lab also does. We initially looked at individual nutrients, how they affect gene expression. Now we are looking at combinations like fructose plus caffeine, fructose plus salt, and then we plan to do like a meal. And then we need to understand how are we going to translate that to a layperson. If I tell a person that, Oh, you have this copy number variants, you need to eat this, the person may not understand anything. So if you are able to convince them that, you have this susceptibility and maybe if you start early to eat properly or do physical activity, you may either delay the onset of a certain disease or may completely prevent it. Why even go for treatments if you can prevent it? Basically, I think the way we part the information, that becomes very important.

Grace Ratley: [00:07:42] Yeah. And so how do you think people are going to get access to this information? Do you think this information will come from screening when people are children or from kits like 23andMe or something along those lines?

Saroja Voruganti: [00:07:56] I like the system of Iceland where they sequence every person, which is okay for a small country, but not for United States. So it may be very difficult, but maybe do something like that. Actually, I would like it to be like when you go to a doctor, they take a blood sample and send it to a lab for lipids or something, lipid panel or something. Similarly, if they can draw blood and send it to a lab or some genetic lab and they can immediately have a set of genotypes which they can do it in a day or two, genotype them and send back their data. And they say, Oh, you have this certain genetic variants, so maybe you should follow this, this, this or something like that. That would be ideal. We are working towards it and I’m quite optimistic that we will be there soon. I don’t know when, but soon. The problem is that we ourselves don’t know what is the role of a lot of genetic variants. So I think first we need to dive deeper into it, understand it, and then come up with a consensus set of like, let’s say these genetic variants are associated with dyslipidemia or these are with diabetes mellitus or something like that. Now we have also complexity with microbiome and other factors, so it’s not easy, but at least if we can take care of 40% of it or 50% of it, it helps.

Grace Ratley: [00:09:24] Definitely. That’s one of the reasons why I really enjoy learning about nutrition. It’s that complexity and figuring out how much of what we look like or how healthy we are can we attribute to our lifestyle or our genetics or our microbiome. I feel nutrition gets a bad rep because it’s so hard to study, it’s so complex. And I feel every study has different criteria and it’s really difficult to look across them and see the same result every time.

Saroja Voruganti: [00:09:56] Yeah, I agree. And also another thing which adds to complexity is we really don’t have good ways of measuring the diet. So all we are doing is food frequency questionnaires and dietary calls, which is so much focused on memory. I don’t remember what I ate yesterday half the time, so I have to think so much. Even if I remember, how much did I eat?

Grace Ratley: [00:10:23] It takes a long time. I’ve used the OneFit app or something like that. Yousit down and it takes 15 to 20 minutes to log everything that you ate. Like I ate five almonds.

Saroja Voruganti: [00:10:35] It’s not easy. So when we do studies, I feel bad for the participants because it’s not easy. I wish we can find something more objective. So a lot of people depend on metabolomics. Metabolomics is a good surrogate, but we need something more.

Grace Ratley: [00:10:52] Yeah. And for our listeners, can you give us a little bit about what metabolomics is?

Saroja Voruganti: [00:10:56] Metabolomics is study of the final breakdown products of metabolism. So like proteins, amino acids will be your metabolites for carbohydrates. You have monosaccharides like fructose glucose, and then for fats they will be fatty acids, different types of fatty acids. But remember that these have all been metabolized and what we have in the blood is what we find. So there will be so much variation in it. But if some person’s diet is steady over a period, maybe metabolomics is actually a good measure till we find something better for dietary intake. For now, they’re very good markers because we don’t have anything else. So we should use that in combination of what you have collected from dietary intake recalls and food frequency questionnaire.

Grace Ratley: [00:11:50] What other scientific tools have been really key to this growth of the field of Nutrigenomics and Nutrigenetics?

Saroja Voruganti: [00:11:58] Mainly sequencing is pretty much within our reach. Earlier it was very expensive and all. Sequencing is getting very cheap so soon I think whole genome sequencing will be very cost effective. I think that is a major thing. And then now we can sequence microbiome. So that is also very interesting. And then metabolomics is getting very cheap. So you can do untargeted metabolomics. If you want to understand general metabolomics, it’s like whole genome sequencing. You do every metabolite measure, every metabolite and see or if you are only interested in a pathway like myself, we go for targeted and look at what is happening in that field. Yeah, but we do need some other like enzyme chemistry, HPLCs or all to add to it. Yeah, I think we are in a much better position now with a lot of nutrition related ones and genetics.

Grace Ratley: [00:12:58] What sort of simplified advice could you give to someone who is looking to improve their health.

Saroja Voruganti: [00:13:03] Based on genetics, I don’t know whether we can advise. So right now I will just say like we always say, do moderation. Don’t stop anything completely, especially macronutrients. Don’t stop completely carbohydrates or fats or protein. The other thing is, of course, portion control, which we always been telling. Even if there’s so much on the plate, try not to eat the whole thing. Try to control your portion for now. A lot of people have shown that Mediterranean diet is actually very good for cardiovascular health. And you have DASH diets, you have specific diets and all. You don’t have to do major changes like in one of your ingredients, you can reduce it. Like salt a little bit less or oil a little bit less, so slowly. And still make a difference. So why not? We are just trying to see how we can keep healthy people healthy, not after they have got hypertension. And we are telling that, oh, follow DASH diet. It’s good for them to follow that diet anyway.

Grace Ratley: [00:14:11] Yeah. And I feel like a lot of people get their health information from doctors who are always on the other end. People go to the doctor when they have something wrong and not so much for prevention.

Saroja Voruganti: [00:14:24] Exactly. And doctors also realizing the importance of nutrition now. And they also refer quite a few people to dietitians. But I agree with you that this is after something has happened. Not everybody, but at least people who are genetically susceptible to certain diseases, maybe they can start early exercising more or eat in a better way or something like that so that they can at least delay and improve their quality of life.

Grace Ratley: [00:14:54] I like this idea of potentially using nutrition as a medicine, as a therapeutic or as a preventative thing.

Saroja Voruganti: [00:15:01] Yeah. And actually, I’m at Nutrition Research Institute. We are big on using food as medicine or food as preventive medicine or something like that. So if we can do it with food, why go for treatment?

Grace Ratley: [00:15:15] Right. For some reason, it’s very difficult to convince people that investing in your health early or investing time and effort into nutrition early can really save so much money and trouble down the line. Why do you think that is?

Saroja Voruganti: [00:15:33] There’s something like my son used to say when he was really small that why is healthy food always so non good tasting or why is tasty food always not healthy? Like if we deep fry something, they are very tasty all the time. If the same thing is baked, it is good, but it’s not the same level. So I think how we present it, how it tastes, smells, it makes a lot of difference on what people eat. That is why Dr. Ammerman, she does something called culinary medicine. Those biscuits are there, Bojangles biscuits. She made like those type of biscuits, but with healthy oils and all. So I really like what she does. It’s she’s trying to present similar food but in a healthier version. So if people take to that, if it smell, taste and everything is good, then I think people will accept it. It’s just that they want to enjoy food. Food is something which is a topic for meetings and parties, culturally festivals. Everybody wants to be healthy. Who doesn’t want to be healthy, but they will also want to enjoy food. I think, yeah, culinary medicine has an important role to play.

[00:16:51] And secondly, I think the time for preparation. We did a study when I was doing my PhD several years ago, that was supposed to be a weight loss study and the mothers ended up. They were all low income mothers. Most of them ended up gaining weight. We were trying to teach them all nice exercises and nice recipes and all, but one lady told me that we work three jobs a day. I have to run from one job to another. Where do I have time to do your exercises or prepare the food? I just pick up something from McDonald’s and take it home. I need to get something for my children. So we have to have foods which taste similar and they have to be affordable to people and should not take too long to prepare because a lot of people don’t have time to prepare them. Once they are not well or once they get some disease, they have no choice. They have to do it. I wish they can do in the earlier part, but nobody will give you a time off because I’m healthy, I want to take some time off. It’ll be counted as vacation or they’ll not give you time off. If I say I’m not well, I want to take one day off. People say, okay, take off a sick leave, but I say I’m healthy, but I still want to take one day off. People will frown on that person. Why they want to take off. It’s my thing that we should also invest in healthy people too, so that they stay healthy.

Grace Ratley: [00:18:16] Perhaps even more so investing in healthy people.

Saroja Voruganti: [00:18:19] Yeah. I know that we have to treat people who are not well. I agree totally. But we also have to invest in healthy people because we don’t want them to move to the dark side or something like that. So yeah, I think we need to do a lot more work, but I’m so excited to be in this field because this is something we can make a difference like you’re born with certain genes, you can’t do anything about it. But how they’re expressed is in our hands and what we eat, like lifestyle factors are in our hands.

Grace Ratley: [00:18:52] Exactly. And I think you brought up an interesting point with the low income women gaining weight and people who are in lower incomes tend to be at higher risk for a lot of these diseases and for having poor diets. And it gets into this accessibility aspect. Do you think that Nutrigenomics will be accessible to people who are in lower income areas or who are diverse in ethnicity or race?

Saroja Voruganti: [00:19:19] Initially, genetic databases were not really representative of various groups, but now there are a lot of groups, especially consortiums, which are trying to get more the genetic databases updated with ethnically different populations. So most of the studies have been done in Caucasians because the study started in Europe and US. So that is how it happened. But now we are trying to push more representation from African-Americans, Asians and Hispanics. So yeah, there is a push, but it will take some time for accessibility. First, we need wide representation in genetic databases. Secondly, we need wide representation in studies so that we know more about each population. I’m studying about a genetically isolated, homogeneous population, which hardly people know about them and they are very hesitant to participate in these type of studies. So we need more representation. I’m originally from Asia, I’m from India originally, but all Asians are clubbed together, but there is so much diversity within Asians. Similarly, Latin Americans. What does that mean? There is so much genetic diversity. Even within Africa, there is so much genetic variation. So we need representation from everybody. So once we have that, then our genetic databases will be very rich and diverse. I know we want to do a personalized nutrition someday, but at least right now our focus is a genetically susceptible group. So if we can divide it into groups or ethnic level, that will be an intermediate step before we actually go to the personalized nutrition.

Grace Ratley: [00:21:01] So how did you get interested in nutrition?

Saroja Voruganti: [00:21:05] Initially, my interest was all biochemistry, and then I had a subject called Food Science, and I took that and I was very interested in how you can prepare different foods, different types of foods. And each food has the way you prepare it, it can taste different. My first experiment, how you whisk that. So if you use different whisks, you’ll get a different types of foams. It was like simplest one and I was so fascinated and I started looking into deeper into it like what is involved in an egg? And then slowly, slowly I got interested in nutrition and nutrition is, after all, biochemistry. Biochemistry and nutrition are so closely linked. So that’s how I got interested in nutrition. And then I said, Oh, this is so cool. And I went into dietetics. But soon I realized that I can’t advise people unless I know more about this subject. What other thing I realized that foods are culturally so deep rooted in communities. It’s so difficult for us to ask them to change. So I thought, okay, why not go get in deeper into it and understand more so when we moved to us, I did not want to pursue dietetics. I wanted to go into research because I wanted to understand more about each nutrient combination of nutrients. And so that’s how I moved to nutrition. But they are all so closely interrelated. So it is you can say nutritional biochemistry that was what I was interested in.

Grace Ratley: [00:22:40] Yeah, that’s what I always tell people. I majored in because then I get less questions about French fries and things.

Saroja Voruganti: [00:22:47] Yeah. So nutritional biochemistry and then I got into nutritional genetics you can say that. And then when the group which I was working with postdoctoral fellow, they were working with several groups like Dr. Comuzzi, like we had datasets from Hispanics, Arabs like Omanis, Arabs, Alaskan natives and American Indians. So I was like, I had so much opportunity to look at each ethnicity and really, really got interested into minority research and diversity research and all. So I have tried to continue that work at NRI or UNC.

Grace Ratley: [00:23:25] It’s very important work and I’m glad that there are people like you out there studying these things and helping not just the average white American, but the average American and the average person in the world. It’s not just an America.

Saroja Voruganti: [00:23:40] Yeah. Anybody like white or anybody. We need to keep healthy people healthy and see that they don’t progress to something. And those who have already progressed have a disease or something, keep them there and not make them progress to become worse or something.

Grace Ratley: [00:24:00] Yeah. Do you do a lot of cooking at home?

Saroja Voruganti: [00:24:03] Uh, I used to. Nowadays I don’t have much time, but I do. And I use very little oil. When my husband prepares the food much, much more tasty because he uses traditional. But my foods are not that tasty. So I keep telling him. And then I saw how much oil he uses. I’m like, okay, I’m not watching when you’re cooking. I can’t.

Grace Ratley: [00:24:29] Yeah, but it’s just not to know how much is in there.

Saroja Voruganti: [00:24:32] Yeah, yeah, yeah, yeah. But that’s why I was telling you, right, that we need to start thinking of how we can prepare the same type of food with less oil, which is I don’t know how, but we have done so much research. We have done so many discoveries, I’m sure we can find something.

Grace Ratley: [00:24:49] Right. I always come back to that. Yes, I know what’s good for me, but I also know what’s really tasty. It’s so hard.

Saroja Voruganti: [00:24:59] Yeah, it is hard.

Grace Ratley: [00:25:00] So last question to young scientists or scientists who are interested in exploring the topic of nutrition, what advice would you give to these people?

Saroja Voruganti: [00:25:12] I’m very passionate about nutrition, so I would encourage people to come in. And also it may apply to any field actually. Don’t expect overnight results. This will resist research. Of course, it will take some time. But what is interesting is that you can make a difference. You can figure out changes and all. So I think I would love to have more people working in nutrition. There’s so much complexity in this field. Just don’t think that it is all about cooking or anything. So just remember that we are like engineers and then doctors are more like mechanics. So someone told me this like I was explaining to them this pathway. And also they told me that, Oh, you are like an engineer, and doctor is like mechanical. So yeah, get deeper into it. And really you have to enjoy the field to be working in this, make this field your own and just deep into it and enjoy. Because not only that there is a lot of complexity and a lot of scope, but this is a field where you can make a difference sooner than later, actually compared to other fields. And luckily most of them do not have side effects if taken in moderation. Nutrition in that way is.

Grace Ratley: [00:26:34] It’s like high benefit, low cost.

Saroja Voruganti: [00:26:36] Exactly. Yeah. And some things I’m very quite excited about the way different ethnic groups have cultures, their own cultural foods, and they all have some scientific significance, even though maybe their ancestors know or do not know about science, we don’t know. But the combinations of a tortilla and beans. There is a variation in India too, and everywhere there is certain variation of that. It makes so much difference because together they have essential all essential amino acids. So at that time, I don’t know if they knew science or not, but that’s an ideal combination, tortilla, wheat flour plus beans. I’m really fascinated by looking at their cultural foods and how well they have all the essential amino acids and they make such a good combination. There’s so much more to learn in nutrition, not just a biochemistry. Biochemistry is part of it, but culturally, socially, psychologically. There’s so many aspects of nutrition. There’s so much to learn. So I encourage the students to come and choose whatever aspect they want to choose. They can choose a social aspect or cultural or genetic aspects, which I’m doing or microbiome aspect. So there is like so much to learn and so much more to investigate. This is like a gold mine.

Grace Ratley: [00:28:03] I agree. Well, thank you so much for coming on our podcast, Saroja. You have a wonderful perspective and so much knowledge and passion about nutrition, and I hope that our listeners can take some of that away with them.

Saroja Voruganti: [00:28:15] It’s my pleasure. It was very nice talking to you.