Visiting Ephesus at #embointegmod25 to look for the missing AlphaFold 3 code

Would you like to come to my lab for a postdoctoral position? Apply for the EPFL AI Centre fellowship: epfl.ch/research/fun.... FYI, I have funding (ERC & Simon's collaboration) also for 1-2 top candidates who will not secure the fellowship. Same for @krzakalaf.bsky.social lab.

💥 We are *also* organizing Machine Learning for Structural Biology @ EurIPS, Copenhagen!! Topics include anything in the ML+bio intersection. Submit your ML+bio short papers! Authors can even present in both locations if they have people in both locations 🚀

Cool-looking work by @piompons.bsky.social @albecazzaniga.bsky.social @fra-cutu.bsky.social, going beyond existing works on sampling different conformational ensembles with AF2.

Recently tested some de-novo minibinders against two targets (thanks Adaptyv!) designed using our open-source design library, `mosaic`; our best method got hit rates of 7/10 and 8/10 and affinities as low as single-digit nanomolar. Wrote up some thoughts here: blog.escalante.bio/minibinder-d...

Posts about research on Bluesky receive substantially more attention than similar posts on X, formerly called Twitter go.nature.com/45Ftiw4

Two exciting openings with us! 🤖🧬🆎🧫💉 - AI Scientist 👉 lnkd.in/eDXHH4E8 - AI Scientist, Drug Creation 👉 lnkd.in/eEvGyaTR You'll work on antibody sequence/structure design, antibody-antigen co-folding, antibody-antigen binding prediction, physics-based methodologies, and more! DMs welcome!

🎉 Excited to share that the last paper of my PhD is now published in PRX Life! We introduce RAG-ESM, a retrieval-augmented framework that makes pretrained protein language models (like ESM2) homology-aware with minimal training cost. 📄 Paper: journals.aps.org/prxlife/abst...

@cyrilmalbranke.bsky.social is 🦋! "ProteomeLM significantly outperforms DCA in recovering experimentally validated interactions. In H. sapiens, ProteomeLM achieves an AUC of 0.83, compared to 0.73 for DCA. Among the top 10 million scored pairs, it recovers 50% of known PPI, versus only 20% for DCA"

A no-miss opportunity to join EPFL. We are a vibrant (had to use it 😉) community working in all branches of basic sciences and engineering, and interdisciplinary studies are strongly encouraged. If you want more information, contact the committee heads, or feel free to DM me. tinyurl.com/bd27ky7p

"When even the most benign activities are subject to capricious disruption and political interference, the luxury of disengagement is no longer a viable option."

AtomWorks is out! Building upon @biotite_python, we built a toolkit for all things biomolecules and trained RF3 with it. All open-source, test it via `pip install atomworks`! AtomWorks: github.com/RosettaCommo... RF3: github.com/RosettaCommo... Paper: tinyurl.com/y2w4z65b 1/6

From social media apps to bank loans and hiring/firing decisions to search engines and shopping sites, are you tired of living a life ruled by opaque algorithms? Pre-order now to Take Back Control on August 5. bit.ly/4idpeGr

Excited to share work with Zhidian Zhang, @milot.bsky.social, @martinsteinegger.bsky.social, and @sokrypton.org biorxiv.org/content/10.1... TLDR: We introduce MSA Pairformer, a 111M parameter protein language model that challenges the scaling paradigm in self-supervised protein language modeling🧵

Our development of machine-learned transferable coarse-grained models in now on Nat Chem! doi.org/10.1038/s415... I am so proud of my group for this work! Particularly first authors Nick Charron, Klara Bonneau, Aldo Pasos-Trejo, Andrea Guljas.

Proved that the diabolical Einstein field equations can be formulated as a well-posed initial value problem, paving the way for huge progress in mathematical general relativity that continues to this date. Goodbye, Madame Choquet-Bruhat

#Postdoc opportunity in #protein #evolution at #EPFL I can highly recommend working with Anne-Florence Bitbol! You'll bounce ideas off a world-class scientist with an awe-inspiring range of interests and a sharp mind—without sacrificing quality of life or a lake view. tinyurl.com/ProteinPostdoc

French scientist denied US entry after officers find phone messages criticizing Trump

Bluesky is the place for science.

Go Sweden! 🇸🇪

🚨 Check out DrugFlow, our new generative model for structure-based drug design. DrugFlow provides an atom-level confidence score for each designed molecule, and can adjust molecular size on the fly! Additional details in thread 🧵 #ICLR2025

Thank you @seanmcarroll.bsky.social: x.com/seanmcarroll.... This has an effect. People want to keep hearing from you and they too will migrate here.

Dear US colleagues, In response to the Trump administration, my university @univ-amu.fr is setting up a new program dedicated to welcoming American scientists wishing to pursue their work in France. There are several M$ in the budget. www.univ-amu.fr/en/public/ac...

Hello American friends here. Please consider making gigantic protests great again. This platform is too small. A centralized app for coordination seems a good idea to me.

POV >Be Ettore Majorana >Discover neutron-proton model of nucleus and predict the neutron first but refuse to publish > Quantize complex scalars but refuse to publish >Predict new kind of fermion >Show possibility of particles with arbitrary spin >Quit physics at 32 >Refuse to explain >Leave ⚛️🧪

Excited to share PoET-2, our next breakthrough in protein language modeling. It represents a fundamental shift in how AI learns from evolutionary sequences. 🧵 1/13

Can we learn protein biology from a language model? In new work led by @liambai.bsky.social and me, we explore how sparse autoencoders can help us understand biology—going from mechanistic interpretability to mechanistic biology.

Back from a short and intense #NeurIPS2024. I enjoyed meeting old friends and catching up with cutting-edge research! I am proud of my team (LIAC at EPFL) with 3 spotlights out of 6 accepted workshop contributions. If you want to join the team, please apply below (deadline December 20th). #chemsky

AlphaFold3 still tends to overfit regions that AF2 multimer predicts to be disordered. Almost all disordered regions become helices in AF3.

Early December, the CASP16 Meeting was held in the Palladium resort in Punta Cana, Dominican Republic. Nick Grishin, myself, and Qian Qong presented the assessment of the monomeric structure predictions, the assembly predictions and of monomers in their oligomeric context. 1/n

Had an awesome time walking among giant ancient trees with a great group of people @neuripsconf.bsky.social. Highly recommended!

I can't begin to describe how life-changing this new project, ShellSage, has been for me over the last few weeks. ShellSage is an LLM that lives in your terminal. It can see what directory you're in, what commands you've typed, what output you got, & your previous AI Q&A's.🧵

Omg this is absolutely incredible 😻 Drew is such a legend, I use his videos in my undergrad teaching all the time, "ok now let me show you how this process realistically looks like"

Crosslinks can sometimes improve AlphaFold3 models. Crosslinking is an experimental method often used as a source of distance restraints in modeling. In AlphaFold3, instead of using restraints, crosslinks can be included as covalent ligands. Thanks to Agnieszka Obarska-Kosinska for the idea.

@torstenschwede.bsky.social @alphafold.bsky.social @capridock.bsky.social In the AF3 paper, massive sampling was shown to improve performance dramatically for antibody–antigen complexes. The AF3 server below justifiably did not sample at this scale. Are there plans to compute this baseline?

Stone conclusion by hearsay from day 1 at #CASP16 all top methods used AF3 somehow. Improvements are made by increased sampling (FFT etc), ranking of models still a challenge. Some improvement on antibodies.

"Before their discovery, a script from 1900 bc in Egypt was the oldest known alphabetic writing; it turned hieroglyphs into alphabetic letters of West Semitic languages."

“On a log-log plot, my grandmother fits on a straight line.” -Physicist Fritz Houtermans There's a lot of truth to this. log-log plots are often abused and can be very misleading 1/5

New paper out in @ScienceMagazine! In 8 studies (multiple platforms, methods, time periods) we find: misinformation evokes more outrage than trustworthy news, when it does it's shared more + ppl are less likely to read before sharing. w/ @killianmcl1 @Klonick @mollycrockett 🧵👇

CAPRI conclusions from the CASP16 assembly session

Assessors’ conclusions of the 3D category (individual protein chains) of #CASP16 🧪

Now I can share external links without the posts being down regulated - I thought it would share this again!:) I have compiled a list of now 100+ companies in the 'TechBio' space into a fully open database for the community. Find here and please share if you like open.substack.com/pub/harrisbi...

Good news: You can post video on Bluesky now! Bad news: Only up to 60s. So here’s a clipped version of my latest video about a paper exploring the landscape of possible molecular glue degradation targets:

#GoodThingsComeToThoseWait Sorry it's taking us a while. Promise that the wait will be well worth it! #OpenFold3 Love all the great work that's already been done #HelixFold3 #Ligo #Chai1 #Protenix #Boltz1

* one year of DESI data is worth 20 years of SDSS! * no evidence yet for gravity behaving any differently to Einstein's general relativity

Bluesky tip for newbies: Bluesky has no real algorithm. Likes here do nothing but show the poster that you enjoyed their content. That kindness is always going to be appreciated by creators, but if you want to help to boost something here, you have to repost it so others see it. 🙂

"It appears that your recommended reviewers are not really your friends, Mr. Bond"

“I’ve noticed I’m not in your starter pack. Any particular reason for that?”

The exodus is real: I lost 200 followers on that other platform and gained 1000 here. Keen to connect! #BioSky #Scisky #Bioinformatics