“Most of the companies we have seen have an AI component to support the discovery or development processes,” Francisco Dopazo, a general partner at Humboldt Fund told TechCrunch recently. But despite becoming quite the buzzword, AI’s apparent ubiquity in biotech isn’t actually driving deal flow or higher valuations. So to get a better idea of how AI is affecting biotech in 2022, we asked six investors to tell us what they look for in a biotech startup today. For Franck Lescure, a partner at Elaia Partners, in biotech, having an AI component isn’t an automatic deal closer. “We do not favor biotech startups with extant AI over those without: Bio-revolution is not only digital. Digital is one tool; the other major tool is the living organism,” he said. VCs are also increasingly looking for what biotech startups can do with AI beyond just R&D, and are wary of companies that use the technology as a marketing tool. “When evaluating ‘AI for drug discovery companies,’ I view AI as a tool,” Shaq Vayda, principal at Lux Capital, told TechCrunch. “Much like how any modern biotech company is using the latest and greatest tools, AI is becoming more and more common as part of biotech workflows. The bigger question for investors is getting a better understanding of what exactly AI is attempting to model and predict.” Also, just because a startup uses AI doesn’t mean it can escape being compared to struggling public biotech comparables. “The public markets are the final arbiters of value, and the valuations coming back to earth this year have begun to flow through to startup funding,” said Sarah Guo, founder of Conviction. “I expect we’ll continue to see some digestion through the next year or two, as many mid-stage companies have built major war chests and don’t yet need to come back to market.” The survey also covers the implications of U.S sanctions on China for startups in the space, considerations for startups thinking of taking capital from government bodies, how to pitch these investors, and more. We spoke with: Robert Mittendorff MD, general partner and head of healthcare, B Capital James Coates, health & human performance principal, Decisive Point Shaq Vayda, principal, Lux Capital Franck Lescure, partner, Elaia Partners Francisco Dopazo, general partner, Humboldt Fund Sarah Guo, founder, Conviction
Robert Mittendorff MD, general partner and head of healthcare, B Capital
The NASDAQ Biotechnology Index peaked in 2021. Have declines in the public-market valuations of biotech companies impacted your investments in the sector? Public market biotechs are dramatically down as interest rates rise and the focus on near-term development outweighs the promise of longer-term results and approvals. As a result, a significant proportion of biotech companies are trading below cash. Given the substantial and positive flow of data in the space, we view market sentiment as overly negative. These valuations have affected private-market rounds’ size, pricing and structure. Private biotechs are considering the reprioritization of their assets — deciding whether to partner second or third assets with strategics, and evaluating structure in tranched financings to reach their fundraising targets. Of the biotech startups you’ve seen lately, how many had an AI component? Do you favor biotech startups with extant AI capabilities over those without? AI has become a very important part of next-generation drug discovery in both the small molecule and biologics spaces. Boston Consulting Group (BCG) partner Chris Meier reported in the March 22 Issue of Nature Reviews Drug Discovery that 24 “AI native” drug discovery companies have a combined 160 disclosed discovery programs. We are many more above this. Recently our own portfolio companies Atomwise and InSilico each inked $1.2 billion deals with Sanofi. Still, the majority of biotechs raising capital are not “AI-enabled.” This isn’t a necessary condition for us, but in many spaces, computational approaches can rapidly improve drug discovery success and speed, at a potentially lower cost. We also see AI being used in the biologics space, although the technology is used there far earlier. AI-enablement doesn’t increase our interest unless the technology is robust, mature and adds value to the platform in a meaningful way. IBM sold Watson Health to private equity in 2022 after investing billions into it. What can biotech startups and investors learn from what could be seen as a cautionary tale? Biotech companies will ultimately be measured largely by their therapeutic pipelines and portfolios rather than by their tech platform. AI for AI’s sake doesn’t hold water anymore. Results, whether in the form of novel therapeutic programs, diagnostic capabilities, or other clinically meaningful outcomes, are necessary. We know quite a few startups are working on AI-assisted drug or protein discovery. Where else can AI play a role in healthtech? AI is a capability, or more accurately described as a set of computational capabilities that can be applied to a set of problems where conventional techniques have demonstrable limitations. AI technology can play a role in biologics, small molecules, and even cell therapy. We have witnessed its application in every aspect of a biopharmaceutical’s business — from discovery, clinical development, and applications in real-world evidence creation to go-to-market motions and post-market patient engagement. AI is not a monolith; as a set of capabilities, the power of learning systems affords benefits to many previously difficult or intractable problems. How commercially viable will personalized medicine be in the next five years? Personalized medicine is already here. See the success in oncology over the last decade, from targeted therapies that are based upon tumor genomics to cell therapies that are N-of-one therapies, where a patient’s own immune cells are engineered to attack the cancer. Personalized medicine as a viable business has already borne out. The question of how far we can go with personalized therapy is the one being answered in the market today. Clearly, many therapies do not need hyper-personalization, but as we learn more about cancer, metabolic disease, and neurological disorders, we are enabled with advancements in biologic and computational science to customize or configure therapies for each patient. Y Combinator welcomed a significant number of healthtech startups in its recent batches. Has YC’s presence had any impact on early-stage valuations? Y Combinator has been a net positive force in driving innovative experimentation at the early stages of company development. Their healthtech cohorts are solid, and their apprentice model works well there. They are still perfecting their approach to projects that focus on biologic science, but I remain optimistic. They have had far less of an effect on valuations for us than the larger momentum firms that recently moved into healthcare over the last few years. How has due diligence in this space changed in 2022? We have welcomed the investment environment of 2022 as both companies and venture investors can diligence each other at a more natural pace. Venture capitalists and founders need time in the process of diligence to understand each other, and the fervent environment of 2021 diminished and, in some ways, attempted to commoditize both. As venture capitalists, we focus on selecting teams and projects that have the highest merit as transformative companies. This exercise takes significant effort and a clear understanding of a number of areas that cannot be accomplished in a day. Diligence is more efficient now than in 2019, but we have returned to a far healthier pace for both founders and VCs. Is Big Pharma interacting more with biotech startups this year than in prior years? When approaching yet-private companies in the space, do the majors favor M&A or corporate venture activity? We are starting to see more deal-related activity pick up, but with a heavy tilt towards business development deals, and some corporate venture activity. Biotech has proven its worth as the engine of innovation for the biopharmaceutical industry, and larger strategics have clear programs for engaging with smaller venture-backed entities. One would imagine given the valuations we are seeing in the venture-backed ecosystem that more M&A would occur given the quality of many of these assets relative to price, but we are at the early stages of that curve. We heard that U.S. sanctions on China could extend to biotech. What impact could this have on AI-enabled biotech startups elsewhere? Clearly, CFIUS continues to have important implications on venture financing across all sectors. Biotech is no different, and there may be more sensitivity moving forward, especially as it relates to advanced technologies, particularly in tech and biology. This may have a modest cooling effect on the pricing of some assets, but I doubt it will affect whether quality companies and teams are funded properly. Should AI-enabled biotech startups take non-dilutive capital from government entities? Why or why not? This is a complex question. If the entity is a U.S. government affiliate, the answer is maybe. For other governments, in particular those outside the U.S. or Europe, it is a more challenging question. Government funding nearly always has conditions of some kind that have to be clearly balanced with the future path of the company. If the funding is from a military source, the implications of dual-use technologies must be considered, and so must be the strategic drift that such funding might encourage. Are you open to cold pitches? How can founders reach you? Yes, but warm pitches are usually better. You likely have someone in your network who is also in mine. My email is [email protected]
James Coates, health and human performance principal, Decisive Point
The NASDAQ Biotechnology Index peaked in 2021. Have declines in public-market valuations of biotech companies impacted startup investment in the sector? Definitely. Going public is a preferred exit strategy for many, and those valuations have just been cut by more than 80%, driving down demand for all but the highest-quality startups. As evidenced by the XBI itself, such cycles are part of the sector. Of the biotech startups you’ve seen lately, how many had an AI component? Do you favor biotech startups with extant AI capabilities over those without? The ubiquity of AI in pitches that I see is striking. It’s hard for a biotech company to convince me they are doing more than just using AI as a component of their R&D (which they probably ought to be!). IBM sold Watson Health to private equity in 2022 after investing billions into it. What can biotech startups and investors learn from what could be seen as a cautionary tale? Commercialization and market expansion are not necessarily immediate downstream consequences of innovation for companies. We know quite a few startups are working on AI-assisted drug or protein discovery. Where else can AI play a role in healthtech? Anything involving data, be it electronic health records or imaging and image-guided procedures. We’re particularly excited by cognitive neuroscience and human performance in this context. Y Combinator welcomed a significant number of healthtech startups in its recent batches. Has YC’s presence had any impact on early-stage valuations? We work closely with many innovative ecosystems in health and life sciences. None of the investments we are most excited about are from YC (at this time). How has due diligence in this space changed in 2022? As I mentioned in my TechCrunch article: cash run-way, non-dilutive capital and market size have reemerged as the key metrics in determining whether or not to invest alongside the science. Should AI-enabled biotech startups take non-dilutive capital from government entities? Why or why not? If it aligns with the commercial trajectory of the company, then yes. If the grant or contract doesn’t align with what the company aims to accomplish, they should not take the funding (unless in life support mode!). Are you open to cold pitches? How can founders reach you?
6 investors discuss why AI is more than just a buzzword in biotech by Anna Heim originally published on TechCrunch