Excited to announce our paper, SHAP zero, has been accepted into NeurIPS 2025! SHAP zero takes the first steps toward extracting biological insights at scale from machine learning models by amortizing the cost of explanations across large-scale biological datasets.

Our new work on the utility of SAEs in low-N protein function prediction and design arxiv.org/abs/2508.185...

Sparse Autoencoders for Low-$N$ Protein Function Prediction and Design

Predicting protein function from amino acid sequence remains a central challenge in data-scarce (low-$N$) regimes, limiting machine learning-guided protein design when only small amounts of assay-labe...

arxiv.org

🚀🚀🚀 Super excited that our Agentic AI for Origins of Life is covered in a story in journal Nature. #ai #astroagents

@celestebiever.bsky.social · last yr.

Team of collaborative AI 'scientists' enlisted to probe the origins of life in the Universe www.nature.com/articles/d41... - with @amiralia.bsky.social and @miquai.bsky.social

Research involves day science and night science, requiring you to become good at both. Day science proceeds in logical steps and specializes in hypothesis testing, but to generate new ideas we pop out to the creative night science world, where leaps among ideas are made possible.

A great conclusion to my Generative and Geometric AI class at GT with several interesting projects and posters. #GenAI

Great read on ruggedness of protein fitness landscapes

Colin Jackson@cjjackson.bsky.social · 2y ago

dealing with the ruggedness of protein fitness landscapes is perhaps the most challenging aspect of protein design and engineering - here we have written a short perspective about it - it's complicated and not perfectly understood but very interesting! arxiv.org/abs/2411.12957

Here’s the updated Computational Biology Starter Pack! Let me know if you'd like to be included. go.bsky.app/QVPoZXp

I must admit that this annotated Nature abstract remains a useful recipe for constructing a summary paragraph. I show it to my students every time we get started.

Research funded by NIH generated $2.46 in economic activity for every $1 of funding in 2023 [a total of $92.9 billion] www.fiercebiotech.com/research/rep...

Report: Every dollar of NIH research funding doubles in economic returns

Research funded by the National Institutes of Health (NIH) generated $2.46 in economic activity for every $1 of funding in 2023, a total of $92.9 billion, according to an

fiercebiotech.com

We strongly suggest that academic publishers and other platforms that host research rapidly implement a Share to Bluesky button for their articles. Here's how: docs.bsky.app/docs/advance... #AcademicSky #HigherEd #Altmetrics

Action Intent Links | Bluesky

Authors, websites, and apps can use action intent links to implement "Share on Bluesky" buttons, or similar in-app actions. Logged-in users will be directed to the corresponding action view in the Blu...

docs.bsky.app

This pack is now up to 67 people, broadly in AI and advanced data for science. Follow along, and forward me others who should be there! 🦋🦋 go.bsky.app/JeFdryY

I'm looking forward to engaging on #chemsky. Missed #chemtwitter!