Great advice from @a16z for those of us building the next generation of biopharma companies. a16z.com/2018/04/30/bui…
As mentioned before, some great advice here! Having been working in the space for 4+ years now, and having fallen into some of these traps ourselves, I thought it would be useful to share some brief twitter reflections from my time leading and building @RecursionPharma!
Excited to share this as @DaphneKoller and @insitro start down this path. So much to be done here, and so many patients waiting for effective therapies - all great minds welcomed! Having buy-in from great investors like @rtnarch should be a strong signal to biopharma!
Pitfall 1/ Starting as a service. Right on to avoid this. Short term revenues, but at the cost of long-term upside. This has been a golden rule for @recursionpharma since day 1.
Pitfall 2-1/ Lacking drug dev exp. Also spot on. Experienced team members like @RecursionMartin on the team, as well as newcomer Kevin Lynch, is essential. Also early on before any institutional capital, supplemented Board with Perry Fell of @SeattleGenetics who is big help
Pitfall 2-2/ These and many others have been essential for us to balance the benefits of our lack of assumptions with the wisdom of experience.
Pitfall 3-1/ Focusing on platform, not assets. I don't see this one as a pitfall, as long as some focus on assets. We are close to our first clinical asset, and have many more in the pipeline. That said, we still spend majority of our effort and $ on platform.
Pitfall 3-2/ This decision mostly because of our belief (and early data), about how many assets it can generate in the long term. Assets today are a means to an end to build the platform, which one day will hopefully create assets at a different scale...
Pitfall 4-1/ Relying only on comp (not Bio). Agree a huge pitfall and one we’ve seen lots of peers make. We’ve had biologists side by side with data scientists since day 1. Today the team is about 1/4 comp/eng, 1/2 bio/dev, and1/4 support/admin/bd, etc. Has been essential balance
Pitfall 4-2/ Also why we have spent majority of our effort and money on building real biological datasets. These datasets enable us to test predictions about mechanism and target, which are now getting quite robust given the >1 petabyte of real bio data we’ve generated so far.
Pitfall 5/ Not picking an indication. Partially agree. Should pick a lead indication, but still OK to work on multiple indications if your platform can show data that it is broadly applicable. We’ve found unexpected traction by working broadly - some pgms have leapfrogged others
Pitfall 6-1/ Not thinking about chemistry. Agree. Our early focus on repurposing enabled us to delay (so as not to split our focus to broadly), but we are now focusing on chemistry. 1st efforts have been combining image data with structure to predict drug tox, to exciting results
Pitfall 6-2/ Excited to announce the start of our in-house chemistry team here in a few weeks as well. Some very experienced folks joining us, and certain we'll see some massive advances, especially on our NCE work.
Pitfall ⅞-1/ Too much in vitro, too little in vivo. Generally agree - I’d argue in vivo is only as good as the model, which often aren’t predictive of the disease. Showing a genetic signal in humans is perhaps better.
Pitfall ⅞-2/ Good data in diverse functional (not pathway) in vitro assays isn’t useless for diseases without good models. Definitely agree that really strong in vivo data in predictive models is gold standard when available and is key milestone for a project.
Pitfall 9/ Thinking about efficacy and not tox. Strongly agree! Also disease context important here to determine appropriate tox. Has been a strong focus of ours for the last yr, and look forward to soon sharing work predicting tox from imaging screening data and chem structure
Pitfall 10/ Not thinking about the clinic. Couldn’t agree more. Should be something you think about from day 1. This is a key reason we have project teams with dev/tox/reg or other clinical insights working on our assets from the day they are called as hits on the screen.
Pitfall 11/ No clear idea of value milestones. Very important lesson for those with a tech background, and those working with tech investors in bio. Value inflections look VERY different here.
Pitfall 12-1/ Not paying for good IP attys. Agree. Easier (and essential) to protect specific discoveries, both methods of use and composition of matter, than to protect platform. Pro-tip - even the best lawyers will give you a volume disc. if you are filing enough patents 😉
Pitfall 12-2/ Doesn't mean there aren't great ways, IP included, to protect platform.
Pitfall 13/ Asking for too much/too little $. Definitely a challenge, especially when $ is flowing (like now). Assess how much $ it will take to get to next big value inflection, then double it as a buffer, then put head down and execute.
Pitfall 14-1/ Signing bad deals (or no deals). This was the biggest learning for our team, as we haven’t had a BD pro until recently. First business negotiation of my life was with worldwide head of BD for a major pharma! Learned a LOT (and still got a good deal).
Pitfall 14-2/ Right lawyer with lots of experience in this space can be an amazing advisor and sharer of norms for those new to this space. We've had an amazing one, and she's taught me so much here!
Pitfall 14-3/ Best advice I’ve heard is to only sign a deal if it is better than the last. Has been good for us, with most recent announced deal here: biocentury.com/bc-extra/compa… . Let's see if next one is better 😀
Pitfall 15-1/ Not building a strong syndicate. Couldn’t agree more. Having great investors that believe in your vision, and aren’t afraid to get in the trenches and build with you, is a blessing. Excited to have the likes of @Lux_Capital, @DCVC, @Mubadala and many others with us!
Pitfall 15-2/ Also helpful to have great angels from the biopharma space as well as from the VC community. Their networks, both in pharma and with investors, have been hugely helpful.
Pitfall 16-1/ Focusing on valuation and not progress. Also agree, with caveat that founders need to know how much money they think they’ll need to make it to vision, and to dilute appropriately along the way. Plan your raises around clear value inflections to help with this.
Pitfall 16-2/ For us, A came after signing 3 pharma deals (product market fit). B came after demonstrating ability to build team, to discover at scale, and to build data packages around assets. C will be when we are clinical stage showing progress/translation of follow-on assets.
Just my $0.02 here. Hope someone finds these helpful as well as the original post. Thanks to @vijaypande, @a16z and others for the great original post! Intersection of tech and bio is an incredible place to work!
