→ 
.
These advances include high-throughput screening (https://en.wikipedia.org/wiki/High-throughput_screening …), compbio/data science/ML, rapid drops in the cost of DNA sequencing, RNA-Seq + transcriptomics, etc...
Show this thread
Pharma R&D productivity continues to decline despite a host of tech advances: "The average cost for [pharma] companies to bring a drug to market has increased to record levels of just under $2 billion, up from $1,188 million in 2010 and $1,539 million in 2016."
-
-
-
The central two problems seem to be: (1) Slow feedback loops -- days (cells in culture) to years (clinical trials). You can't speed these up (unless you figure out how to simulate them). (2) Our (relatively) poor/nascent understanding of human biology.
Show this thread -
On #2, as is well known, presumed mechanisms of action mostly don't predict correct clinical outcomes, because of the complexity of human biology. On this front, this FDA report looking at a bunch of drugs that failed in Phase II is illuminating: https://www.fda.gov/downloads/AboutFDA/ReportsManualsForms/Reports/UCM535780.pdf …pic.twitter.com/7FYV0qbckg
Show this thread -
All of this is to say: I’m seeing a *ton* of VC funding of “AI to cure cancer / discover drugs” type companies. But if they’re not solving the two problems I identify above, it’s not clear what fundamental advance would make somebody bullish on them.
Show this thread -
The goal is to make drugs that will be effective *in live humans*. If you don’t have a way of speeding up the feedback loop, or if you don’t have some fundamental advance in biology/knowledge, then how are you going to be 10x better than big pharma at this?
Show this thread -
Moreover — getting results with AI requires both fast feedback loops (gaming!) and proprietary data. How (if at all) are these startups getting/generating data that is proprietary and that gives them some kind of competitive advantage?
Show this thread -
I’m curious to hear from people who are more bullish on this than I am. Until we fix the above problems, it seems to me that drug discovery will not accelerate in the way that more internet-ish fields are, and I would be bearish on the average biotech startup.
Show this thread -
I also glossed over explaining the 2 fundamental problems above properly, because this is Twitter, but would be happy to write an essay on this if there’s interest
Show this thread -
It wouldn’t surprise me if R&D productivity were still in stagnation in 20 years. One implication of this for people who want quicker results (which is most tech people) is that “prevention” is a more promising vector for health than “curing”.
Show this thread
End of conversation
New conversation -
-
-
You should talk to my friend Elizabeth
Thanks. Twitter will use this to make your timeline better. UndoUndo
-
Loading seems to be taking a while.
Twitter may be over capacity or experiencing a momentary hiccup. Try again or visit Twitter Status for more information.