Real precision medicine is starting with the patient: Harness patient biology and AI to enable biomarker-driven personalized medicine

Emily Le:
Hello everyone. I'm Emily Le from Molecular Medicine Tri-Conference. I'm here with two experts from the world of precision medicine, and I'm very pleased to have them in this podcast today. I have Niven Narain, Cofounder, President and CEO at BERG Health, who will be giving a keynote presentation at the Precision Medicine Meeting as part of the Tri-Con in San Francisco on March 11, 2019, and Niven Narain, General Partner at New Ventures Funds who will be chairing at the AI to support clinical decision-making session at the Molecular Diagnostic Strategy and Outlook Meeting at Tri-Con. Harry will also be teaching a short course on Commercialization Boot Camp: Manual for Success in Molecular Diagnostics, and he will be at Tri-Con for his book signing on commercializing novel IVDs. I'm very delighted that Harry has agreed to join us for this podcast today, and he will be interviewing Niven today about his talk at Tri-Con on the topic of Harnessing Patient Biology and Artificial Intelligence to Enable a Biomarker-Driven Personalized Medicine Approach. Thank you both for joining us today.

Harry Glorikian:
Thank you very much for having us.

Niven Narain:
Thank you, Emily. It's my honor.

Harry Glorikian:
Niven, it's really good to talk again. You're always pushing the boundaries on this conversation, and we've had it for many, many years, but I think for those people that don't know much about your proprietary platform, I mean I've had the pleasure of sort of digging into it quite a bit over the years, can you give a brief understanding of what it is and why this talk is critical to this group of people that's coming?

Niven Narain:
Niven Narain: Sure, Harry. I'd be happy to. First, mantra and philosophy over the years is that we needed to really take a back-to-biology approach to how we firstly understand patients. In so doing as we create the products into developing therapeutics and biomarkers, what we feel really brings together a mantra of real precision medicine is starting with the patient. BERG has developed the platform over the years that harnesses the power of clinically-annotated tissue samples that are then exposed to a multi-omic expression profile that allows us to glean insights into what is ongoing at certain time points in the case of longitudinal sampling. Then once we have a level of expression data, we then either recreate the disease-relevant stimuli in the wet lab, so those patient-derived cells then undergo a layer of perturbation biology where the omics is done all over again in those environments. What we're able to do is be able to appreciate over certain time points, over certain environments, what is really ongoing within the disease environment and microenvironment. Then we correlate the expression data, functional data, to the clinical EMRs, demographic and real world evidence. Then all these datasets are then subjected to a Bayesian artificial intelligence system.

About eight years ago we really pioneered this effort to bring together biology and AI so that we could really achieve two things. One, add more precision to the process of drug development and diagnostic development, but also be able to engage more reproducibility, because I think the two biggest issues in the industry is really a lack of precision and understanding of the biology, vis-a-vis drug and diagnostic development and the reproducibility of the biology around drug development and clinical trial development. Really, put simply, this combination of biology and math that BERG has been able to harness under a fully integrated platform which is referred to as interrogative biology.

Harry Glorikian:
Niven, we've been doing measurements of populations and looking at expression profiles for ages. I don't even want to talk about how long I've been doing this, but I really do believe, and you and I have discussed, that it's really the computational advances that have made some of what you're doing possible. It's still in its nascent stage and moving pretty rapidly. Every time I pick up something to read, there's something new going on in this space. How do you look at what I want to call old approaches versus where we're headed now from a computational perspective where I'm talking to people and I'm hearing before when we build a lab, 80% of it was wet lab, 20% was IT, and now maybe it's the reverse or 70% IT and 30% wet lab.

Niven Narain:
Yeah, I think that's a fair point, Harry. I think, obviously, it's really hard to argue that in the past three to five years that we've had an exponential expansion of not only the awareness education and use cases around informatics, different types of AI, but I have to harken back to an old statement by the Nobel Laureate, Albert Szent-Gyorgyi who said, "True discovery is looking at what everyone else has looked at and seeing what no one else has seen." We have looked at different types of expression data, but I think a lot of it has been so genomically or transcript-driven. One of the unique things that BERG has done is brought together different data topology so that we're taking a deeper slice into the biological narrative of those patients. I agree with you. I think that the only way that we were able to do that, because you're creating terabytes of data on patients. On each individual patient sample, I think that the computational power to super compute capability, the ability to whether you're using a new role or Bayesian methodology, is the ability for us to really bring together these data sets, is really only now I think in the past few years capable of bringing together just the raw multitude of data capture. I agree with you that the infrastructure and architecture perspective has really enabled this.

Harry Glorikian:
Niven, let me ask you a question just so people get calibrated. How do you compare yourself to, let's say the status quo from I don't know how much money was invested versus what would normally be invested, how many people would be involved versus what would normally be involved, but some sort of metrics that someone can understand what this technology approach provides or enables?

Niven Narain:
Niven Narain: Yeah, if we go by the DiMasi numbers just using time, on average it takes about three to five years to come up with a validated drug candidate, and BERG has been able to do that in less than 18 months. If you look at a time perspective just at the front end of biology, being able to come up with a high quality target, validating that target in the wet lab, coming up with a drug candidate, whether it be small molecule or biologic, and the ability to create a druggable therapeutic that would then be subjected to pre-IND testing, our ability through this technology to at least decrease that lead time in half, and I would say from a cost perspective, at least half.

Adding the efficiencies to the front end of the drug development process, I think, has been really our sweet spot. As you make your foray into clinical development, the ability to, since you understand the mechanism of action, you understand the patient population that you'd like to go after. You have at least a general idea of the molecular markers that are correlated to a clinical phenotype, the way that you design and the way you conduct your clinical trials look and feel different, that the types of conversations that you're having with regulatory agencies are very different. The ability to have dependent on the clinical indication, the ability to have some inkling or some ability to understand if the drug is working much earlier is so crucial because compared to not having the capacity to look at a molecular marker, a biomarker, and having to wait on the end of the trial or an interim readout, this could be a huge deal for small companies that are VC backed, has a certain runway or even for a larger company that has to make bigger decisions on potential naval expansion or post market access decisions. The ability to design these clinical trials cannot be underappreciated.

Harry Glorikian:
Yeah, I know and you and I both know that you have the capability to do this. These people aren't falling from the sky that has the knowledge to be able to do this. You laugh, but I've spent many, many, many hours and days interviewing and talking to people to understand the real ones from the pseudo experts. Just in a quick nutshell for those people that are going to show up to your talk, what would you like them to take away from it?

Niven Narain:
What are the global considerations that we should keep in mind going forward? You're working with these great U.S. organizations, but are there any global considerations we should keep in mind as you go forward in your research?

Harry Glorikian:
Yeah, I think that this talk is going to be focused on going through how we harness that power of patient biology and AI to enable biomarker-driven personalized medicine, so meaning that using the biomarkers not only as a readout but using the biomarkers as a vehicle and an agent to design clinical trials much differently, but also to validate patient responses much earlier. Also for diseases, we're seeing a real massive increase in the incidence of age-related diseases. When you look at fields like neurodegeneration or certain types of cancers that progress so quickly, prognostic markers are so important. I'm going to focus my talk on how we can use biomarkers not only as a readout to the actual phenotype, but use it really as tools to drive more causal efficiency in drug development.

Niven Narain:
Thank you for being part of the Molecular Medicine Tri-Conference and its Companion Diagnostics and Clinical Biomarkers program. We are looking forward to meeting you in San Francisco. What are you planning to accomplish by attending and presenting at the meeting?

Harry Glorikian
That's great. Well, for all of you that are going to join us and Tri-Molecular, both Niven and I look forward to meeting you, talking to you and hoping that you take away some incredible learnings from the event. Niven, thank you so much for joining me in this conversation.

Niven Narain:
Thank you, Harry. I look forward to seeing you in San Francisco and participating in this very timely and important conversation.

Mark Strong:
Thank you. I look forward to it too, and hopefully the weather's a little warmer there than it is in DC right now.

Emily Le:
Emily Le: Thank you both so much for your time and insights today. We just heard from Niven Narain, General Partner at New Ventures Funds and Niven Narain, Cofounder, President and CEO at BERG Health. They will be speaking at the Precision Medicine Meeting and the Molecular Diagnostic Strategy and Outlook Meeting as part of Tri-Con this March in 2019 in San Francisco. If you'd like to hear them in person, go to for registration information and enter the key code podcast.

I'm Emily Le. Thank you for listening.

Register Now
March 26-27, 2024

AI in Precision Medicine

Implementing Precision Medicine

At-Home & Point-of-Care Diagnostics

Liquid Biopsy

Spatial Biology

March 27-28, 2024

AI in Diagnostics

Diagnostics Market Access

Infectious Disease Diagnostics

Multi-Cancer Early Detection

Single-Cell Multiomics