Vasilis Ntranos (@vntranos)

Pinned post

Excited to share that our latest work building on ESM is now published in @NatureMethods:

A single, sequence-only protein language model achieves state-of-the-art variant effect prediction, surpassing hybrid approaches that use MSA, 3D structure, or population genetics data.

nature.com/articles/s4159…

www.nature.com Compressing the collective knowledge of ESM into a single protein language model Nature Methods - A co-distillation framework is used to iteratively adapt sequence-only protein language models for high-accuracy variant effect prediction, without the need for additional...

Vasilis Ntranos @vntranos May 27

A tour de force from the ESM team @biohub! Huge progress in protein structure prediction from *single* sequence!

Alex Rives Verified account @alexrives May 27

Today we're announcing ESMFold2, an open scientific engine to power prediction, design, and discovery across protein biology.

The new model delivers state of the art performance on protein interactions, especially antibodies, a critical modality for therapeutics.

We have designed and validated miniprotein binders and single chain antibodies across five therapeutic targets that are important in cancer and immunology. We are seeing very high success rates, and affinities at levels consistent with therapeutic activity.

We’re also releasing an atlas of 6.8 billion proteins, and 1.1 billion predicted structures.

ESMFold2 is built on a state of the art language model that has been trained on billions of protein sequences.

A world model of protein biology emerges through language modeling.

We’ve used the techniques of mechanistic interpretability developed to understand large language models to understand the concepts ESM uses to represent proteins.

The model’s representation space has a compositional organization of features across scales, levels of complexity, and abstraction, that reflects and mirrors the understanding of protein biology developed through a century of empirical science.

This understanding emerges without prior knowledge, just from language modeling of protein sequences.

Language models are becoming a powerful substrate to understand and program biology.

The design of protein interactions is one of the most fundamental problems in biophysics, and has critical implications for the discovery of new medicines. A simple gradient based search with the model was able to discover high-affinity protein binders.

I'm excited by the potential this has to accelerate basic science and the understanding of proteins. And especially for the new avenues it opens up for therapeutic design and medicine.

View quoted post on X

Vasilis Ntranos @vntranos Apr 6

Great work showing that protein language model embeddings can sometimes be indistinguishable from noise. This aligns very well with our observation in nature.com/articles/s4159… that motivated VESM / co-distillation.

Indeed, big knowledge gaps exist even within models of the **same family**, with almost identical architecture and trained on similar data: E.g., ESM2 and ESM1b each fail to detect well-conserved domains that the other recognizes as mutationally sensitive.

Nature Methods @naturemethods Apr 6

A model-agnostic empirical framework measures the uncertainty associated with protein embeddings and assesses their biological relevance to improve model reliability. @y_bromberg @rpkarans

nature.com/articles/s4159…
View quoted post on X

Vasilis Ntranos @vntranos Aug 11, 2023

Excited that our paper is now published in Nature Genetics: nature.com/articles/s4158…

Overall we show that protein language models are an effective, accurate and general approach to predicting variant effects. 1/n

www.nature.com Genome-wide prediction of disease variant effects with a deep protein language model Nature Genetics - A modified framework leveraging a protein language model (ESM1b) is used to predict all possible 450 million missense variant effects in the human genome and shows potential for...

Vasilis Ntranos @vntranos Apr 21, 2023

👉UPDATE: You can now use #CellTypeWriter without a #GPT4 api-key. We've set up a #free version so you can try it out on your #data. ✨This will be active for a few days or until funds run out : ) Have fun! Github link: github.com/ntranoslab/cel…

github.com GitHub - ntranoslab/celltypewriter Contribute to ntranoslab/celltypewriter development by creating an account on GitHub.

Vasilis Ntranos Verified account @vntranos Apr 17, 2023

Introducing #CellTypeWriter, a user-friendly tool to explore and analyze #scRNAseq data using an interactive chat interface powered by #GPT4. Check it out on github: github.com/ntranoslab/cel…

View quoted post on X

Vasilis Ntranos

Latest posts 5