Kyle Ferchen

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…

@ossia @davidjmalan Thanks @davidjmalan for being part of such a valuable online learning resource! I would want to learn his thoughts on how the CS job market will change with LLMs becoming more capable of autonomous coding. How does he see the the CS50 curriculum adapting to this?

Tomorrow I'm interviewing Harvard CS50 instructor @davidjmalan for the freeCodeCamp podcast. What should I ask him about?

New “discrete state” model of hematopoiesis published in Nature Immunology. rdcu.be/eL8oO We propose a hierarchical model of hematopoiesis where stable “discrete states” serve as key regulatory nodes. With @nsalomonis @morris_lab @BioLegend @CytekBio @NIH @NIDDKgov and tweetless NHLBI and Harinder Singh.

pyInfinityFlow is maintained on GitHub (github.com/KyleFerchen/py…) and provides both command-line execution and a Python API, which are documented using readthedocs: pyinfinityflow.readthedocs.io (6/6)

pyInfinityFlow generated Infinity Flow objects can be saved as FCS files and loaded into #flowjo, or other Flow Cytometry analysis software, to carry out in-silico gating side-by-side with UMAP coordinates to visualize the enrichment of target cell populations: (5/6)

We are excited to announce the publication of our new #Bioinformatics application! #pyInfinityFlow, enables combination of overlapping Flow Cytometry panels, unlocking the potential to profile more features across millions of cells: academic.oup.com/bioinformatics… (1/6)

edfdfd codepen.io/KyleFerchen/pe…