🔥The 9th Round of Easy Loan, Earn $40 Reward is in progress❗️ ⏰ Promotion Period: January 15th - Feburary 15th, 2025 👉 Register now and check more details at gate.io/campaigns/358

Evidently, attention - on data with long range dependencies such as MSAs - really is all you need to solve protein structure prediction. #AlphaFold2 I wonder why more #CASP14 teams didn't try to incorporate this architectural advancement (except e.g. Kiharalab, tFold & RaptorX).

Google’s deep-learning program for determining the 3D shapes of proteins stands to transform biology, say scientists. nature.com/articles/d4158…

"Es tut mir wirklich von Herzen leid, aber..." – Bundeskanzlerin Merkel mit einem emotionalen Corona-Appell im #Bundestag. #Generaldebatte

Last year we presented #AlphaFold v2 which predicts 3D structures of proteins down to atomic accuracy. Today we’re proud to share the methods in nature w/open source code. Excited to see the research this enables. More very soon! bit.ly/alphafoldmetho… bit.ly/alphafoldgithub

EMBL and @DeepMind have partnered – a breakthrough for science. Together, we're providing a treasure trove of protein structure predictions powered by #AlphaFold to herald a new era for #AI-enabled biology. embl.org/news/science/a…

Now that the #alphafold hype has completely died down (ha!), I've written a new blog post on the AF2 method paper: moalquraishi.wordpress.com/2021/07/25/the…. This is a technical deep-dive into aspects of AF2 that I find most surprising/innovative and of relevance to broader biomolecular modeling.

Cracking the regulatory code: We have genomes for 1000s of species, but ENCODE only for 2 – what do we do? Natural language models have shown that syntax and semantics can be learned from text alone. Can we do the same for genomes?⬇️ biorxiv.org/content/10.110…

Is binarization of scATAC-seq data necessary?Quantitative modeling of scATAC-seq data performs better at latent space learning and at the same time preserves quantitative information of accessibility. Join my talk at #ISMBECCB2023 Tue at 14:50 & Poster B98! HiTSeq 2024

Have you ever wondered what the genome looks like through the eyes of a DNA language model? In our newest preprint we use DNA LMs to study nucleotide dependencies in the genome, revealing functional elements, characterizing variants and evaluating DNA LMs tinyurl.com/6wbwjaf4

Excited to present “scooby”, which models multi-omic profiles (scRNA-seq coverage & scATAC-seq insertions) directly from 500 kb DNA sequence at single-cell resolution. This was a fantastic collaboration co-led with Johannes Hingerl. biorxiv.org/content/10.110…

Want to aggregate rare variants into a single gene score while capturing diverse functional effects → Check out DeepRVAT, my first paper lead with Brian Clarke now out in Nature Genetics. DeepRVAT scores can be plugged into any testing framework to discover new disease genes.

From computational models of the splicing code to regulatory mechanisms and therapeutic implications go.nature.com/4gPju63 #Review by Charlotte Capitanchik 🌙, Oscar Wilkins, Nils Wagner, Julien gagneurlab & Jernej ule_lab Helmholtz Munich | @HelmholtzMunich TU München The Francis Crick Institute King's College London

The Borzoi manuscript is now out in Nature Genetics: doi.org/10.1038/s41588… Borzoi predicts RNA-seq profiles in many tissues & cell types from DNA sequence as its only input. With it, we can score the impact of genetic variants on a number of gene-regulatory functions. 1/

Can DNA sequence models predict mutations affecting human traits? We introduce TraitGym, a curated benchmark of causal regulatory variants for 113 Mendelian & 83 complex traits, and evaluate functional genomics and DNA language models. Joint work w/ gokcen and Yun S. Song 🧵👇

1/ DNA sequence models like Borzoi predict gene expression and variant effects across 1000s of tissues — but what if your data comes from a custom experiment? David Kelley Johannes Linder and I propose a lightweight solution: parameter-efficient fine-tuning (PEFT). biorxiv.org/content/10.110…