In a new Perspective published today in @GENBiotechJrnl, John Chodera (@jchodera) and I give an optimistic view on the future of neural network potentials (NNPs) for enabling advanced small molecule discovery and design. Summary Thread Below! 🧵.

Under or Over? . By 2050, radiography will be completely taken over by cell-free ‘omics for monitoring solid tumor progression.

Congratulations to the @MonteRosaTx team for having their work featured on the cover of @ScienceMagazine!. We (@_DimensionCap) are tremendously excited about the future of molecular glue degraders (MGDs) and induced proximity writ large. Even in 2025, we've only scratched the.

As AI/ML approaches cross over from the scaled manipulation of bits to the physical world of atoms, it makes all the sense to me that they start with small molecule therapeutics. The $/atom economics are the highest. Positioning 20 atoms could be worth >$1B/year.

One of the best parts of my role is having the privilege of being surrounded by passionate and brilliant scientists working on extremely difficult problems. The @chaidiscovery team falls into that category unequivocally. I was pretty blown away by this data as well as the.

RT @AriWagen: Chai Discovery (@joshim5 @chaidiscovery) just released Chai-2, a model for "zero-shot antibody design in a 24-well plate.". T….

RT @bauer_lesavage: We had an excellent time attending our first #ASCO conference this year - so much so that we decided to write up our ke….

- Executing on that vision with smart clinical development and formulation.- Priming the market for a successful launch. Nowadays, Biohaven is a much bigger, more diverse standalone company and Nurtec ODT is important enough for a pop icon to talk about on TV.

At the end of the day, Biohaven differentiated itself by:. - Excelling / having conviction in a target pathway.- Being savvy dealmakers and taking advantage of a non-consensus buying opportunity.- Having excellent TPP-driven strategy for how to inject value w/ minimal cost.

At the end, they had enough data to not only get the ODT formulation approved, but also enough to get the dual label, which they won in 2020 and 2021, respectively. They were off to the races. Nurtec ODT generated >$350M in the six months leading up to the acquisition.

These trials enrolled >4,000 people and evaluated Nurtec against placebo as well as Nurtec ODT against placebo ~ all in the acute setting. Then they ran another Phase II/III for prevention.

Their hypothesis was that ultra-rapid symptom relief (<15 minutes) over placebo would matter to the mass market. They were right. Next, Biohaven decided to run four Phase III trials in parallel. The 301/302/303 trials were a parallel pivotal/confirmatory grouping.

What Biohaven did next was very clever. They added value through formulation and the first ever dual-label for acute and preventive migraine. Biohaven licensed Catalent's Zydis ODT technology for Nurtec (rimegepant), which they felt was critical.

But beauty is in the eye of the beholder. Biohaven in-licensed BMS's drug and later paid $50M upfront in exchange for a better royalty structure. Experts in the CGRP pathway, Biohaven seemingly felt compelled by the BMS molecule and its ability to avoid DILI class effects.

Later on, Merck had an oral CGRP antagonist meet its primary endpoints in a Phase 3 study, but it had some hepatoxicity issues, leading to its voluntary termination. This seemingly spooked Bristol-Myers Squib, who shelved their CGRP drug after a Phase IIb study.

Triptans act indirectly on the CGRP pathway, validating the mechanism. But triptans also have a high rate of migraine recurrence and carry cardiotoxicity warnings. Boehringer developed an IV-infused CGRP drug in 2004, but it didn't catch on.

From day one, Biohaven was all about the CGRP pathway, having spun out of a world class Yale neuropsychology lab. Around the 2010s, several companies were working on CGRP inhibitors to outcompete the contemporary migraine standard of care: triptans.

Even so, Biohaven's story with Nurtec ODT is fascinating. Nurtec ODT is a CGRP antagonist. It suppresses pain and blocks pathologic arterial dilation in the brain. We've known about CGRP biology for >40 years. So, why did Biohaven capture most of the value here?.

Neuro is a particularly difficult therapeutic area (TA). It has the 4th lowest likelihood of approval the 3rd longest development time. Granted - migraine is a special case because it's easier to enroll and quicker to readout than neurodegenerative diseases.

Pfizer owns Nurtec ODT. It's projected to do about $1.5 billion in sales this year. Pfizer acquired Nurtec ODT in 2022 when they bought Biohaven (which they subsequently spun back out). I was surprised that Biohaven commercialized a neuro asset independently.