We did 370 experiments to discover that protein language models primarily learn structure and won't scale for protein function prediction. We need new pretraining tasks! Work led by Francesca-Zhoufan Li @ ICLR'25 with Ava Amini Yisong Yue Alex Lu See Alex's thread + the paper for more!

"Enhancing the efficiency of protein language models with minimal wet-lab data through few-shot learning" arxiv.org/abs/2402.02004 1/

Happy to share the final version of our work describing computational filters for selecting functional ML-generated enzymes in Nature Biotechnology! This was a distributed and remote collab with @TrichomeDoctor Xiaozhi Fu and other members of Aleksej Zelezniak🇺🇦's group. Global Science!

🎉 Pilot Protein Engineering Tournament results are in! See benchmarking results for enzyme design models from in silico💻 to in vitro🧫 & the champion🏆Data, submissions, scripts & more!📊 Discover all the details here: alignbio.org/the-tournament… #ProteinEngineeringTournament2023

Active machine-learning guided evolution of protein function! Ravi Lal and Jay engineered two properties simultaneously in a new-to-nature #enzyme. #directedevolution. Useful method!

Great work by Kadina Johnston and her team--here is a complete enzyme 4-site dataset for #ML4proteins, folks! #enzyme.

Where our carbon comes from will define our climate future. We developed a synthetic carbon assimilation pathway with new to nature enzymes, enabled by #cellfree synthetic biology. #sustainability Ashty Karim biorxiv.org/content/10.110…

Computational design of serine hydrolases Institute for Protein Design 🚀 New preprints from David Baker! 🚀 1. Groundbreaking study uses advanced computational design to create novel serine hydrolases with unprecedented catalytic efficiencies. These new enzymes could revolutionize

De novo enzyme design!!! | Amazing work from Anna, Sam David Juergens and all the co-authors! Computational design of serine hydrolases biorxiv.org/content/10.110…

Thank you to the Food and Health @ UC Davis for the invitation to share our work, and a big thank you to @joanna_c_chiu Justin Siegel and @AAbrieux! I especially loved the honey tasting, what a treat and great addition to discussions on nutrition and circadian biology!

A couple years ago, Judith Amores convinced me that olfaction was cool, then we convinced Seyone Chithrananda to join us for an internship (he had options!). Hopeful that we can do even cooler things together! Code: github.com/microsoft/olfa… Preprint: biorxiv.org/content/10.110…

Stay tuned 🎧

We are thrilled to share that Dean H. Rao Unnava has been named Dean of the Year by Poets&Quants, a leading global business school media company. JOHNABYRNE highlights Dean Unnava's innovation and leadership over eight years at #UCDavis: bit.ly/3BteXGl

The wait for trainable SoTA protein folding model is over! We are happy to introduce Protenix (Protein + X Molecules!), our Open Source reproduction of AF3, including inference code, dataset, weights, and training pipeline. Check our code: github.com/bytedance/Prot… 🧵 1/N

Train a protein structure flow-matching model that can be conditioned by atomic coordinates and then use it to design enzymes based on active site geometry specified by density functional theory.

”David Baker, Demis Hassabis and John Jumper, your ground-breaking work in computational protein design and protein structure prediction has revolutionised these fields. It has opened up completely new possibilities to design proteins that have never been seen before, and we now

Feels like we summited Mount Everest, celebrating David Baker's Nobel Prize. De novo enzyme design was thought to be impossible by many... look where we are now! Honored to be invited to this celebration, to share my expertise, and reunite with friends. Onward & upward! Arzeda

🚀 Collaborative, AI-Ready datasets would revolutionize life sciences and catalyze progress. Government funding plays a critical role in making this a reality. 📖 Read our policy memo here: buff.ly/4hbH78p

✨ Newly launched! Our Special Report: Fermentation: An Ancient Solution to Modern Challenges, is now online. Read the report to explore how traditional fermentation techniques can support sustainable and equitable food systems, especially in LMICs. 🌍 This report highlights