Tim O'Donnell

Anyone willing to donate GPU time to re-train MHCflurry?

MHCflurry 2.2.0rc2 is on PyPI: pypi.org/project/mhcflu… Try it out and let us know if you spot any problems in our transition from TensorFlow to PyTorch

I suppose I am coming back to this platform... bsky never quite hit critical mass for me

Thanks to @iskander 's robot army for rewriting mhcflurry in torch! Available now as "pip install mhcflurry==2.2.0rc1" github.com/openvax/mhcflu…

I'm rebuilding AlphaFold2 from scratch in pure PyTorch. No frameworks on top of PyTorch. No copy-paste from DeepMind's repo. Just nn.Linear, einsum, and the 60-page supplementary paper. The project is called minAlphaFold2, inspired by Karpathy's minGPT. The idea is simple: AlphaFold2 is one of the most important neural networks ever built, and there should be a version of it that a single person can sit down and read end-to-end in an afternoon. Where it stands today: - ~3,500 lines across 9 modules - Full forward pass works: input embedding → Evoformer → Structure Module → all-atom 3D coordinates - Every loss function from the paper (FAPE, torsion angles, pLDDT, distogram, structural violations) - Recycling, templates, extra MSA stack, ensemble averaging — all implemented - 50 tests passing - Every module maps 1-to-1 to a numbered algorithm in the AF2 supplement The Structure Module was the most satisfying part to build. Invariant Point Attention is genuinely beautiful — it does attention in 3D space using local reference frames so the whole thing is SE(3)-equivariant, and the math fits in about 150 lines of PyTorch. What's next: - Build the data pipeline (PDB structures + MSA features) - Write the training loop - Train on a small set of proteins and see what happens The repo is public. If you've ever wanted to understand how AlphaFold2 actually works at the level of individual tensor operations, this is meant for you. Repo: github.com/ChrisHayduk/mi…

new preprint alert! tl;dr we made a global tokenizer for proteins

@iskander Nice! This has been such a long running to do...

Claude Code did like 95% of the MHCflurry TensorFlow to PyTorch port in about a day but there are some persistent numerical differences/bugs that it's been churning for a week, playing trial-and-error whack-a-mole coding like a junior dev. Codex seems to be one-shotting the fix

Re-posting Nick Boyd's Mosaic blogpost after his incredible results in the Nipah G competition. blog.escalante.bio/180-lines-of-c…

🚀 Just released: Protein Hunter on GitHub! github.com/yehlincho/Prot… Now supports Boltz and Chai with more models coming soon! Use it to: 1️⃣ Design binders from scratch 2️⃣ Optimize your own designs 🔗 Boltz: shorturl.at/0s5Ih 🔗 Chai: shorturl.at/NqrHZ

OpenFold3-preview (OF3p) is out: a sneak peek of our AF3-based structure prediction model. Our aim for OF3 is full AF3-parity for every modality. We now believe we have a clear path towards this goal and are releasing OF3p to enable building in the OF3 ecosystem. More👇

@iskander @eryney_ok @owl_posting Totally agree. @seromics want to screen some de novo designed binders?

@eryney_ok @owl_posting I’d probably use a protein array like this: pmc.ncbi.nlm.nih.gov/articles/PMC60… And if you wanted more fine grained but less sensitive mapping then something like PhIPseq It would also be reassuring to see flow or IHC of cells with and without targets to assess non-protein stickiness

At some point earlier this year I told @owl_posting that I didn’t think binder design was solved. I was wrong.

I’ve been testing BoltzGen a bit recently and while I haven’t done any experimental testing yet, the quality of the software is very clear. It installs, runs, logs everything, has tons of options. Very excited to test out the designs irl!

Excited to release BoltzGen which brings SOTA folding performance to binder design! The best part of this project has been collaborating with many leading biologists who tested BoltzGen at an unprecedented scale, showing success on many novel targets and pushing its limits! 🧵..

🧵1/ We @tahoe_ai just published a new post on the Tahoe blog—a story of how we used Tahoe-100M, the world’s largest drug-perturbed single-cell dataset, to find compounds that upregulate MHC-I and make tumors more visible to the immune system. Here’s how 🧬🔍👇

Protein templates are finally supported in Boltz-2! Next up: updating BoltzDesign1 to use Boltz2 for template-based motif scaffolding and more

Excited to unveil Boltz-2, our new model capable not only of predicting structures but also binding affinities! Boltz-2 is the first AI model to approach the performance of FEP simulations while being more than 1000x faster! All open-sourced under MIT license! A thread… 🤗🚀

Despite a lot of cool work in the last few years on high throughput TCR/epitope mapping assays, there's still only ~30 pMHCs where we have at least 100 known binding paired-chain TCRs. Where is all the data? Is it locked up in companies?

I moved to Cambridge MA! If you're around and would like to meet up let me know. Looking forward to getting to know the community here

@iskander I wasn't actually - didn't realize that! I was influenced by Lambda Labs prices 🙂

@timodonnell Were you also influenced by the Bengio lab?

Opinions of Bizon Tech for desktop ML / biofx workstations? (ever since I saw MILA students crushing it with 1TB memory / multi-GPU workstations in the mid 2010s I've wanted to ditch academic clusters, might be time to finally do it)