Akshay

Research @ Yale | NYU.
Work on Machine Learning | BioML | Drug Discovery

[Akshay @SpinQED]

Biology is a representation learning problem waiting to be solved

[Ilija Lichkovski @carnot_cyclist]

Science is the act of capturing pockets of computational reducibility

[Akshay @SpinQED]

@TimothyDuignan @navvye Curious why you think NNPs are clearly the long-term solution here. Mutation effects seem to depend heavily on evolutionary constraints captured by PLMs as well as structural and thermodynamic effects.

[✒️ @Literariium]

— Marcus Aurelius

[Pranam Chatterjee @pranamanam]

Input a protein sequence + the residue positions to target → potent motif-specific binders in seconds! 🪄 Introducing our newest version of moPPIt! Experimentally validated for domain- and IDR-specific binding, as well as receptor inhibition and CAR-T cell design! 🧫

📜: https://t.co/d2ei8r04FP
🤗: https://t.co/ZbV51OrN1i

🧵👇

[Akshay @SpinQED]

@meganium1236251 @alec_helbling That's not true. We found evidence that REPA works for Image Editing models as well.

[Elliot Hershberg @ElliotHershberg]

On AI Infrastructure in Biology

https://t.co/T0cw61hwLX

Billions of dollars are flowing into AI investments for biotech.

A lot of this money is going to new platform companies with aspirations to develop drug pipelines. People debate the magnitude of impact this will have, but it's becoming consensus AI will play a central role in the future of drug discovery.

But some of this investment is going to companies selling technology, not drugs. This thesis is met with much more skepticism. Historically, it's been challenging to build big businesses in biotech without a drug pipeline.

In a new essay, I tell three stories of how our industry might evolve to support the development of new AI infrastructure businesses.

Better tools can raise our ambitions for what can be built with biology.

[Akshay @SpinQED]

@mkurman88 This is so amazing!

[Akshay @SpinQED]

@pranamanam Congratulations Pranam!

[Akshay @SpinQED]

@rishabh16_ Hey I’m at MIT for the Boltz Hackathon and the MoML Conference. Would love to meetup!

[Akshay @SpinQED]

@parmita Will you be there for the M-boltz hackathon on 20th at MIT? if yes, would love to meet up?

[Jay Yang @Jayyanginspires]

No one talks about how peaceful life gets when you’re 'locked in'. 12+ hour days feel effortless. Week days and weekends blend together. You wake up, work, and by the time you check the clock it’s time to sleep again. Runners call it the high. Athletes call it the zone. Writers call it flow. Whatever it is, it’s beautiful.

[Z Fellows @zfellows]

Sam Altman: “It is hard to be wildly successful at anything you aren’t obsessed with.”

[Akshay @SpinQED]

@MirunaCretu2 @PrescientDesign Would love to chat about BioML. Your DMs are closed though!

[Akshay @SpinQED]

@rishabh16_ Would love to meetup!

[Akshay @SpinQED]

@Andrei_Tarkhov Those models were useful in other ways even if they sucked compared to linear baselines?

[Akshay @SpinQED]

@TeresaNHuang I always keep missing out on the best talks at Flatiron. Need a better way to keep track lol

[Akshay @SpinQED]

@candrianillborn From where is this? 😅

[Akshay @SpinQED]

@charliermarsh uv is so incredible! Deeply regret not switching over sooner.

[Akshay @SpinQED]

@Waymo Awesome!