Had a great time today giving the Chemistry Department seminar and meeting folks Fordham University! Thanks to Joshua Schrier for the invitation and warm welcome!

The twitter algo now shows me 45% Kevin K. Yang 楊凱筌 and 55% Wolf Conservation Center 🐺 .. I am basically ok with this:: food, parenting, good papers and a wolf once and a while

Day 2 of #MLCB2022 kicking off now! Industry panel hosted by David Kelley and featuring Ziv Bar-Joseph (Sanofi), Richard Bonneau (@PrescientDesign/Genentech), and Safiye Celik (Recursion) starting now! youtube.com/watch?v=vKjMZ_…

Welcome to #NeurIPS2022, the world’s largest AI conference! 🌐 DeepMind will be presenting over 40 research papers, hosting ten demos and supporting workshops with partners. 👋 Come visit us and say hi! 📍Booth #609, Hall G

my opening talk at #NeurIPS2022 had possibly the biggest audience i've ever spoken to (a few thousand). thanks (or blame) to the organizers Sanmi Koyejo Shakir Mohamed Kyunghyun Cho Alice Oh alekh agarwal Danielle Belgrave for inviting a philosopher and amateur #ML enthusiast to speak!

Diffusion models excel at protein design including binder design! Check out Joe’s thread to see how we made a pM binder straight off the computer!

The meetup will be at the Rusty Nail at 8 pm *today* (12/01)! 1100 Constance St, New Orleans, LA 70130

I managed to leave Madison airport just before 10am, park in some random lot (very far away from ORD terminal 1) at 11:40, take a slow shuttle bus, grab tram, get to Terminal 1, go through security, run the length of T1 🦕, and make my flight that started boarding at 11:50. 😓💪🏻

Read the full now publishers paper from Alexandra Siegel, Evgenii Nikitin, Pablo Barberá, Joanna Sterling, Bethany Pullen, Richard Bonneau, Jonathan Nagler, and Joshua Tucker 5/5 csmapnyu.org/research/trump…

inspired by some of our work on AF2Rank (showing AF can be used to denoise template inputs) and RFdiffusion, I tried hacking AlphaFold to be a diffusion model for generating backbones. 😎

I am silly excited to share this massive project led by Kotaro Tsuboyama. He developed a new approach to measure folding stability for almost a million protein domains in parallel and we are releasing all the data!!

Wow !

We’ll be presenting at #LMRL2022 tomorrow alongside our partners at Genentech! Catch our poster from 2-3pm MT & our live presentation at 7-8am MT to learn about our framework for building & benchmarking Maps of #Biology. Learning Meaningful Representations of Life Read the paper: bit.ly/3VkAXrX

Our work on applying network inference to neurons in fly, science.org/doi/10.1126/sc… w/ Claudia Skok Gibbs and Neset Özel (NYU CDS and NYU Biology) The Desplan lab's absolute mastery of the Drosophila visual system meets single cell genomics meets our network inference methods

Beautiful work on elucidating the combinatorial TF code for 🪰 neuron specification (from NYU Genomics dream team of Desplan & Richard Bonneau)!

Stephen Ra Learning Meaningful Representations of Life Richard Bonneau Prescient Design Genentech Great talk by Richard Bonneau ! I cast my vote for it to reach the masses if the Learning Meaningful Representations of Life orgs agree; excellent overview of what's possible when ML is deeply integrated with experiments

Take a lexicon match the domain sentences Contrastive loss over them (per batch train to represent similarly sentences that fit the same lexicon and differently otherwise) underline.io/events/342/ses… arxiv.org/abs/2210.15172 Patrick Y. Wu Richard Bonneau Jonathan Nagler Joshua Tucker

Thank you to everyone who turned out (both in person and virtually) to #MLSB last week! MLSB has become a thriving community at #NeurIPS. We had >80 submissions, >200 in-person attendees, and posters on every inch of the walls. (1/4)

New paper on TCR recognition with ML methods as core model from Harri Lähdesmäki. It was very cool of Emmi and Harri to loop Vlad and I into this! Kiitos ! academic.oup.com/bioinformatics…

Last Tweet for a while, account dormant