Molly Gibson, Co-Founder of Generate Biomedicines, on Innovating at the Intersection of AI/ML and Biotech
Looking for more life sciences stories? Check us out at Spotify, Apple, Soundcloud, or LinkedIn
In this episode, we talked to Molly Gibson, Principal at Flagship Pioneering and Co-Founder and Chief Innovation Officer of Generate Biomedicines. Flagship Pioneering is the venture capital firm behind Moderna, and incubates and funds disruptive, cutting-edge biotechnology. Generate Biomedicines is a biotech company that leverages machine intelligence to computationally generate de-novo (i.e., does-not-exist-in-nature) protein therapeutics. Generate recently closed a $370M(!) Series B financing in November this year, highlighting the rapidly burgeoning space of computationally-powered drug discovery.
Prior to Generate, Molly co-founded Tessera Therapeutics and led the computational biology team at Kaleido Biosciences. Molly’s work has resulted in multiple pending patents and publications in both Science and Nature, and in 2020, she was featured in Endpoint News’ 20 under 40 list in biopharma. Molly received a B.S. in computer science from Truman State University, and a Ph.D. in computational and systems biology from Washington University in Saint Louis.
In this podcast, Molly and I discussed:
- The evaluation of commercialization potential for early-stage biotechnologies.
- The intersection of machine learning, computation, and biotechnology: what makes it such an exciting space — and why we’re seeing it boom now.
- The differentiation of Generate Biomedicines, and the implications of its bioplatform on their business model, commercialization strategy, and broader biopharma landscape.
- Advice for early-career business-minded professionals seeking to grow in biotech.
Start to 5:45: In pursuit of curiosity
- On discovering her passion for computer science: Molly’s organic curiosity about how things work in the world around her led her to explore diverse disciplines growing up, including pursuing several majors until she took her first computer science class. The course sparked an intense passion in the engineering principles of teaching computers to imitate human thinking, and to go beyond.
- On pivoting her curiosity to biology, and transforming it into venture creation: Molly realized she wanted to direct her computational passion towards human health. Despite never taking a biology course, she began her PhD in studying how to model the microbiome — the collection of microorganisms in our body — using computational methods! Not satisfied with staying in academics and searching for greater impact, Molly transitioned to Flagship where she continues to seek new ways to transform computational biology discoveries into patient impact.
5:45 to 12:00: Growing early-stage biotechnologies
- On assessing the commercialization potential of early-stage tech: In approaching a new biotechnology, Molly first considers its breadth of impact and its value proposition. From the onset, she believes we must clearly understand the path to how the technology creates value. With the right innovation, the value can actually transform the parameters of the problem it solves, like finding “ghost glasses” that let you see new ways of seeing how the world could work!
“Think value proposition. If we are able to make this biology into biotechnology, something we can engineer and use, then what’s the value to the world? What will we have changed? That’s your most direct route [to commercialization]”
- On the unique model of the Flagship Pioneering ecosystem: Flagship’s venture creation model enables the creation of “bio-platforms”, with deep pipelines that are de-risked via complementary technologies across the Flagship ecosystem. Flagship incubates companies through a team of internal scientists and entrepreneurs with a shared ‘building’ mindset across investments. This model allows incubated companies to quickly tap into a vast expert network, while centralizing operations (e.g., IP, HR, Finance) that enable early-stage biotechs to wholly focus on R&D.
12:00 to 16:35: Intersection of AI/ML, computation, and biotech
- On why this space is so exciting: Traditional drug discovery is limited by empirical trial-and-error approaches, limiting process control and resulting in low success rates, low R&D productivity, and exorbitant R&D costs. By applying engineering principles to biology, we can add reproducibility and predictability to R&D. Machines can tackle biological complexity in ways the human mind can’t, and engineer new biological systems that improve upon Nature, instead of being limited by it!
“Biology is just too complex for the human brain to fully understand…we can now use a new language, the language of AI, and apply that to biology to start to truly engineer new biological systems. It’s paradigm-shifting — moving from empirical discovery process to one where we can control the outcome we want”
- On why only now this space is rapidly expanding: Given its interdisciplinary nature, this space required the confluence of several key advances across domains. (A) Being able to read and write DNA efficiently and at low cost, (B) creating new ways to build up DNA into protein, and © building up algorithms & compute power were all critical to propel this space forward.
16:35 to 21:45: Deep-dive on Generate Biomedicines
- On what differentiates Generate Biomedicines: The company focuses on ‘generative biology’: generate entirely novel biological systems that don’t exist in nature. Consider developing a car engine: they start with blueprints, define what they want it to do, and design and create the engine from scratch (except this process is applied to proteins!). Generate uses ML to learn the fundamental principles of proteins and create novel proteins with optimized properties.
“We have this incredible advantage, where we can learn from all of Nature, but then generate things that Nature has never seen before”
- On commercialization potential: Generate believes it can develop therapeutics better, faster, cheaper. The process can rapidly generate therapeutic candidates, with computationally optimized properties, that can then be built and tested in the lab, the results of which can be used to teach the algorithms to generate better candidates. This ‘build-test-learn’ cycle aims to streamline R&D timelines, improve probability of success, and cut R&D costs from lower failure rates.
“This capability will change the speed and specificity by which we create drugs. We can generate hundreds of diverse candidates [near] instantaneously that we then build and test in the lab… and computationally optimize. It changes the timeline by which you discover drugs, as well as the efficacy and potency of the therapeutic candidates. It’s a real game-changer”
- On commercialization strategy — figuring out “where to go first”: Given the broad optionality of the platform, selecting where to prioritize remains a key challenge. Furthermore, partnerships with Big Pharma will be critical given the number of therapeutic candidates that Generate’s platform can develop, which holds implications on which indications and modalities they explore first.
21:45 to 25:50: Impact of AI/ML bioplatforms on biopharma
- On the disruption of ML-driven platforms like Generate Biomedicines on Big Pharma: Traditionally, Molly believes the axis of value creation resided in later-stage drug discovery. However, with ‘more shots on goal’ from these AI/ML platforms, there’s potential for the value creation and innovation step to shift to earlier-stage drug discovery, radically disrupting the biopharma landscape.
- On when we will begin seeing the results of AI/ML’s impact on biotech: She believes we will see the effects of AI and drug development in the next 2–3 years, followed by ‘multiple evolutions’ of the technological impact on drug discovery for decades to come.
“The challenge of treating, curing, and preventing disease is so massive that any time we get a foothold in one area, we will see the next mountain to climb. [AI/ML] doesn’t let us start from scratch every time we climb a new mountain… [it] should let us start-up halfway up the mountain before we start climbing.”
25:50 to End: Advice to next-gen biotech business leaders
- Believe in yourself: Molly strongly believes we must have the confidence to put ourselves out there, and be unafraid to be wrong! With both passion and humility, we should have the confidence to test our ideas and maximize our potential.
