Jason Yim

Proteina: Scaling Flow-based Protein Structure Generative Models Recently, diffusion- and flow-based generative models of protein structures have emerged as a powerful tool for de novo protein design. Here, we develop *Proteina*, a new large-scale flow-based...

I'll be at ICLR. Come check out our generative modeling work! Reach out if you want to chat.

Proteina: openreview.net/forum?id=TVQ...
Protcomposer: openreview.net/forum?id=0ct...
Generator matching: openreview.net/forum?id=RuP...

Coming soon! We were juggling finishing both our PhD and the paper 🥲.

I really enjoyed seeing how protein generation models scale with more data and weights. Congrats to Nvidia and the core contributors for this amazing work!

See our preprint on a new computational protein design benchmark!

Congrats Simon! I enjoyed our time together while starting protein design at DeepMind. Excited to see what you build. Consider joining latent labs if you're interested in ML and bio!

More work needs to be done on RL-like fine tuning to incorporate auxiliary objectives during diffusion training. Current ways of using auxiliary losses can be greatly improved I think.

Chimera: Accurate retrosynthesis prediction by ensembling models with diverse inductive biases Planning and conducting chemical syntheses remains a major bottleneck in the discovery of functional small molecules, and prevents fully leveraging generative AI for molecular inverse design. While ea...

new preprint on chemical synthesis ML models

- showing how to combine multiple models in a principled way
- modern Transformers + GNN to featurize chemical reaction:
- new insights in where the models shine
+ bonus: find the quirky named reaction!

Feedback welcome!

arxiv.org/abs/2412.05269

Structure-based drug design with equivariant diffusion models - Nature Computational Science This work applies diffusion models to conditional molecule generation and shows how they can be used to tackle various structure-based drug design problems

After two years, our paper on generative models for structure-based drug design is finally out in @natcomputsci.bsky.social

www.nature.com/articles/s43...

Excited this is out! I learned a lot interning with Andrew Foong, @franknoe.bsky.social, and the team. Check out Frank's thread and the preprint to learn more.

🚨Our Machine Learning Force Field Mattersim is now available! 🚨

Check it out here 👇
msft.it/6013oBZLt

The force field is designed to be used on a vast range of temperatures and pressures, try it yourself :)

Feedback and suggestions are very welcome!

👋

In a gratuitous attempt to acquire more followers myself 😁, I've made a start on a "starter pack". Hopefully as more people from 🐦 make it over to 🦋, we can extend this a bit. Suggestions welcome!

I've noticed not all accounts seem to be eligible to be added, anyone know what's up with that? 🤔

👋