How Artificial Intelligence Can Accelerate Drug Discovery and Bridge the Translational Gap with Jo Varshney of VeriSIM Life

 

 

The average cost of developing a new drug is estimated at over $2 billion. Roughly 10% of pre-clinical drugs ever make it to human trials. And each year, $50 billion is “misspent” in the trial & error method of research and development. An AI-enabled “decision engine” could help researchers find cures to rare diseases, bridge the translational gap between animal models to human systems, and move faster in drug development reducing associated R&D costs. 

Today on the show we chat with Jo Varshney, the founder of VeriSIM Life, about the AI-enabled technology that is cutting the time for drug development in half. Searching for a drug candidate for bone cancer led Dr. Varshney to tackle the challenges of better predicting outcomes, improving methods for drug discovery, and building models closer to human systems.

About Jo Varshney and VeriSIM Life
Dr. Jo Varshney is a multi-disciplinary and celebrated leader in veterinary & human health development. She is the founder and CEO of VeriSIM Life, which is building AI-enabled biosimulation technology to empower researchers to predict how potential drug candidates will interact in animal models. Dr. Varshney holds a Ph.D. in Comparative Oncology and Genomics from the University of Minnesota, a Master's Degree in Translational Pathobiology & Bioinformatics from Penn State University, and a Doctorate of Veterinary Medicine from the College of Veterinary Science & Animal Husbandry. She serves as a Scientific Advisor at SVAI, a San Francisco-based non-profit organization promoting education and cooperative research at the intersection of computational and life sciences.

Links: 

VeriSIM Life

LinkedIn | Jo Varshney

Genomics Hackathon

Wikipedia | Drug Discovery

Wikipedia | The Translational Gap

Qualio Website

Previous episodes: https://www.qualio.com/from-lab-to-launch-podcast

Apply to be on the show: https://forms.gle/uUH2YtCFxJHrVGeL8

Music by keldez

Transcript

Transcript is automatically generated. Please kindly excuse any grammatical and spelling errors.   

Kelly Stanton: 

Hey everyone. Thanks for joining us on From Lab to Launch today. I'm Kelly from Qualio and I'm your host. If you haven't already please subscribe and give us a review on Apple or Spotify. Also, if you want to be on the show, fill out the application, linked in the show notes. We've had a lot of people reach out from all over the world and we love connecting with you. It's been a lot of fun introducing you to innovators in life sciences. And today we have Dr. Jo Varshney, founder and CEO of VeriSIM Life. She was referred to us by a guest on another episode, Tonja Dowe from Debiopharm Innovation Fund, and we're finally able to have her. Joe is a multidisciplinary and celebrated leader in veterinary and human health development. She founded VeriSIM Life in 2017 to accelerate drug development with AI enabled technology, essentially cutting in half the time it takes for drug development. We'll link to more in the show notes, but we're excited to go deeper on this modernization and the challenges of drug development today. All right, let's bring her in. Hi, Jo. Thanks for joining us today.

Dr. Jo Varshney: 

Hi Kelly. Thank you very much for the opportunity to speak on behalf of VeriSIM Life. It's one of my favorites topics to talk about all day long. And I'm very glad that our calendars finally aligned. So I'm truly excited to, you know, chime in whenever. And if any kind of questions come about drug development. Yeah.

Kelly Stanton: 

Exciting. Well, so tell us how things got started at VeriSIM Life then.

Dr. Jo Varshney: 

Yeah, frustration. I think if I have to put everything in one word, it was truly frustrating to see how slow the progress I was making during my PhD in comparative oncology and genomics in developing a drug candidate against bone cancer, which is a quite common in large peer of dogs and not so common in pediatric population. So mine was to find a key marker that I could target for both of these species. And unfortunately, that process made me realize that there is something, what we call in drug development, translational gap, which is trying to translate the sciences in R and D, that is, doing animal testing, doing lab testing into human systems because ultimately we want the drugs to work in humans, but unfortunately, if I tell you that Tylenol was tested in rat and got extrapolated to humans based on rat systems, you would take a moment to realize, oh wow. So, my dosage was determined by a rat and there was no real closer system that could mimic what I am feeling. Maybe a challenge that may not be so obvious, but it becomes very obvious when it comes to drug development. So long story short, this frustration of trying to find a model that represents closely to humans system really led me to build VeriSIM Life and in addition to that, there was a Google Hackathon that happened, I would say, around early 2017, that really opened my eyes about how computational methods could enable solutions that are existing, but not that obvious through just using the human intelligence and combining human intelligence and machine intelligence could open new doors that have been hidden. And that scalability was something really made a seredipitous route for me to find opportunities to develop this technology. And now, which is, above and beyond my expertise, of course, that really led to the foundation.

Kelly Stanton: 

That's exciting. So obviously trying to mimic those human models. That is a huge challenge for sure. What are some of the benefits of using your platform versus the traditional process?

Dr. Jo Varshney: 

Fantastic question. So I'll give it a best analogy. Here is over $50 billion is misspent in R&D. And what that misspend means is comes down to trial and error method. So there are several hypotheses, you know, researchers have to test out in different lab systems, as well as animal models, before they can come to a good conclusion of what would be a good candidate to move forward. Now, imagine taking this trial and error methods and building a unified framework, which takes this existing knowledge and predicts what could be all those potential combinations and outcomes in half the time, which is what we do. So then you're not misspending that time and money that you may need to test out all the hypotheses in real systems. And instead use our platform to have a guided decision-making and then invest the right resource and time into the right drug candidate and move faster, which is ultimately to the patients.

Kelly Stanton: 

That is a wow. That, that is a huge challenge to overcome for sure. And that was 50 billion with a B that you said, right?

Dr. Jo Varshney: 

Yes, that's right. Yep.

Kelly Stanton: 

And people wonder why their drugs are so expensive.

Dr. Jo Varshney: 

Yes. I mean, this is one reason why we see big labels, because the rarer the condition is, the more expensive it gets because the more R&D spend they have to do, and we are really trying to reduce that burden, R&D burden. And de-risking those clinical decisions at the early stage of drug development. So instead, you know, at the late stage, when you've spent millions of dollars into the candidates, and then you realize, "oh, I could have done a better formulation strategy. I could have used a different dosing strategy. I could have tested different types of dosing based on the human, types of different human." We call it patient stratification. Those are great answers to get earlier on in the drug develop than at the late stage, when, you know, you've invested all, all the resources to get to that answer.

Kelly Stanton: 

Definitely. Tell me a little bit about some of the unexpected challenges then that you've faced with this technology.

Dr. Jo Varshney: 

Oh, we'll be sitting here for hours. The biggest thing to the challenge is the adoption and education. So, you know, it's very obvious for experts in the field who have spent 20, 30 years developing drugs, you know, you have a high throughput bunch of drug assets or candidates that are coming into the pipeline and doing the conventional way of R and D. And then I linked for IND application to get FDA approval. Imagine having them understand a better, a new way to educate on like, you know, if we combine human and machine intelligence, your time and resources could be spent in a much better way. And basically we're asking them to rethink what they know best. And I think that is a big challenge in itself. And I would say the other challenge is identifying, folks who have interdisciplinary or cross functional learnings, as I mentioned, like most people in drug development have great expertise in very specific areas, which what ends up happening with that is it creates knowledge silos because, you know, when you're great in something you're really going focused and going into the deep knowledge of trenches of understanding that disease, that expertise, and really getting that experience. But then there are other aspects that are needed and unfortunately the knowledge silos, kind of, really impact that understanding of okay, how let's computational methods could work with experimental systems with animal systems and ultimately the human systems, for example. So one of the big things we've done is we've created this platform: a newer form of methods of doing drug development.

Kelly Stanton: 

How do you see this technology evolving in the next five to 10 years?

Dr. Jo Varshney: 

So right now we are very focused on translation because we see this as a huge gap from, how drug candidates move from R and D to human. And once we have established a successful, several successful partnerships, which we are very close to doing so, we want to really enable furthering the clinical trials because there are a lot of different challenges in clinical trials that are also part of the translational aspects, but they are being taken into account post the acceptance by the FDA and say like, "okay, now you got to do your phase one, phase two, phase three." But want to combine those challenges back into a translational engine to enable furthering of accuracy of the decision-making engine that we have and improving the translation index by taking into account all the other clinical challenges that we are not taking into account

Kelly Stanton: 

Working with some maybe existing data sets and existing trials and sort of retrospectively analyzing them to see what you come up with?

Dr. Jo Varshney: 

Yeah, absolutely. So right now we have humans as one of our models, virtual models alongside different virtual animal systems to make the predictions. But then we foresee taking into account very specific personalization between different human systems. How do we identify a female population? It's a very big question. In drug development, females are highly underrepresented, for several reasons. We want to really answer those hard questions through our system, based on all of the centralized learning our system goes through on a regular basis.

Kelly Stanton: 

That's exciting and that is a huge problem to solve for sure. Women and female modeling and the complexities there,

Dr. Jo Varshney: 

Minorities. Yeah. And then different kinds of genomics. I mean, this is why it's more of a ten-year plan because, you know, we can't solve everything at the moment, but based on our generalized framework, we would be able to start tapping into these specific, but local challenges within a patient population. And really being able to answer more key questions, how different patient population would respond to the same drug, what could be done different so that the patient enrollment process could become more efficient. The regulatory bodies would be able to understand a little bit more about why a trial needs to be set up in a different way. Then, you know, a more standard way and things like that.

Kelly Stanton: 

So a topic of the day, of course, being COVID 19 pandemic, did you guys experience an increased demand or acceptance around the technology as a result of that?

Dr. Jo Varshney: 

Yeah, absolutely. I think COVID fortunately or unfortunately, well, unfortunately for the world, but fortunately for us became a good representation of how much you can do with computational resources. They're not doing with computational resources. Many of the clients that we have today were struggling to get to the lab to do simple animal experiments or doing more R and D work. And then they started really looking for alternative solutions. Like, "Hey, are there ways we could speed up our translatability? You know, we can raise more money, get more answers, move faster" because everyone who was developing drugs, the focus is to get to the patients. Right? So they started really looking into, technology like ours, but also fortunately, we got government funding to really look for answers for treatments, from drugs that are already existing in the markets and see whether or not that could target Covid in patients where we're seeing higher mortality, because those are the vulnerable patients. And we want to make sure that if we can provide some guidance around that, we'll help those patients in, you know, die who are in dire need for that.

Kelly Stanton: 

Definitely. As a founder and a leader of this company, what would you say are your biggest challenges or opportunities this year?

Dr. Jo Varshney: 

Yeah. So we have successfully shown our technology can enable better decision-making and we have some successful case studies where we have really reduced the time and cost for clients. And now we are really using our engine and asking deeper questions. What if we ask our decision-making engine to identify key candidates for rare disease programs. For diseases where there's no cure, you know, with that said, we have a new asset partnerships that are coming up with some top research hospitals, as well as our own asset programs. And interestingly with one of our own assets, we have been able to receive orphan drug designation in less than a year time, which in a conventional method, it would take three years plus to get to the same place, especially for rare disease. So it's giving us the confidence, but now we are at a place where we want to leverage this confidence into more proving that they can bring more drugs to the market faster and really walk the talk that we are. We are doing it with other clients. Now we want to also do it with ourselves

Kelly Stanton: 

Exciting too. So if you could go back and tell yourself something at the start of your career, what would that be?

Dr. Jo Varshney: 

Oh, I wish I would've started this sooner. That would be, I think the biggest advice. I mean, I don't want to mention this as just from the woman perspective, but you know, being the minority and, being a veterinarian and a PhD first or assigned this first before an entrepreneur, the path to discovering and running a business is very unclear. Right? They don't teach you that. And I just always knew that I'm attracted to hard problems. I always want to solve complex thing. That's like just one of the ways I function. It took me a while to build that confidence, to get to a place where I'm I'm okay, when I'm not okay. I'm able to speak my emotions and being able to fully speak about a vision, which is not existing. And now it's finally there and. Yeah, I'm truly grateful and honored with all our supporters, advisors, investors, and board members who have really supported me from just an idea of where, you know, we were trying to prove things and it was just like, "oh, how do we scale a virtual mouse?" to a decision-making engine and getting to the translational index. So it's quite a journey and I'm really happy I was able to do this.

Kelly Stanton: 

That is an exciting journey to hear about. Where can people go to learn more follow along and connect with you?

Dr. Jo Varshney: 

Well, there are several ways folks can reach out. We have our website: VeriSIMLife.com, info@verisimlife.com is always active. Please reach out there. You can directly reach out to me. It's my first name, last name and VeriSIMLife.com. And we are always eager to learn and understand translational challenges of for different folks. And, you know, one of the things I mentioned it to our clients and I'll mention it, in general is: we are not the experts. You are the experts. We enable that expertise using our platform and help you guide and make the best decisions and provide you what potential probability of success you have with your drug program. So I hope you will consider us and work with us and move faster.

Kelly Stanton: 

That's exciting. Thank you so much for your time today, Jo. It's been a real pleasure.

Dr. Jo Varshney: 

Thank you, Kelly, very much. I appreciate it. And thank you everyone.