🔥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

i enjoyed writing this with owl ! come read about cool science out of the baker lab

PREPRINT OUT ✨ "hack" your microscope to work as a 3D printer! print micrometer sized features at cm scale. a method for easy and cheap microfabrication, with a focus on biological applications. THREAD ↓ biorxiv.org/content/10.110…

Isolated in just 3 years! (Much faster than the 12 years in our previous study). We cultured two strains from 'Hodarcaheales', proposed to be closest to our archaeal ancestors. biorxiv.org/content/10.110…

Rapid and accurate prediction of protein homo-oligomer symmetry using Seq2Symm Nature Communications 1/ This study presents Seq2Symm, a deep learning model fine-tuned from ESM2 for predicting the symmetry of homo-oligomeric protein assemblies, outperforming template-based and prior deep

🔥 🔥 🔥 🔥 shapiq Shapley interactions for Machine learning Forget about SHAP it was more holes than Swiss cheese and relies on unrealistic assumptions #xai

Our MIT class “6.S184: Introduction to Flow Matching and Diffusion Models” is now available on YouTube! We teach state-of-the-art generative AI algorithms for images, videos, proteins, etc. together with the mathematical tools to understand them. diffusion.csail.mit.edu (1/4)

PyPropel: A Python-Based Tool for Efficiently Processing and Characterizing Protein Data 1/ PyPropel is a Python-based computational tool designed to streamline large-scale protein sequence analysis, focusing on efficient data preprocessing, feature generation, and

I taught a grad course on AI Agents at UCSD CSE this past quarter. All lecture slides, homeworks & course projects are now open sourced! I provide a grounding going from Classical Planning & Simulations -> RL Control -> LLMs and how to put it all together pearls-lab.github.io/ai-agents-cour…

📢📢 "Proteina: Scaling Flow-based Protein Structure Generative Models" #ICLR2025 (Oral Presentation) 🔥 Project page: research.nvidia.com/labs/genair/pr… 📜 Paper: arxiv.org/abs/2503.00710 🛠️ Code and weights: github.com/NVIDIA-Digital… 🧵Details in thread... (1/n)

No More Tears with Matryoshka🪆 & Let's Embrace Sparsity 🚀🚀 With sparse autoencoders + sparse contrastive learning, we can compress SOTA text/image/multimodal embedding models from 2k/4k dimensions to 16 active dimensions with: ~100x faster at large-scale retrieval, minimal

A Model-Centric Review of Deep Learning for Protein Design　arxiv.org/abs/2502.19173 AF3, Boltz-1, Chai-1やBindCraft, Pinal, ESM3など、モデルごとにまとめた総説 protenixがない

iqtree/iqtree3 github.com/iqtree/iqtree3

The code & camera-ready version of our #ICLR2025 paper on "Multi-domain Distribution Learning for De Novo Drug Design" are now available 📚 Paper: openreview.net/forum?id=g3VCI… 💻 Code: github.com/LPDI-EPFL/Drug… (1/4)

> sparse MoE autoregressive pLM > compute optimal scaling law > 1.5T tokens of external/internal data (Profluent Protein Atlas v1) > functional alignment via IRPO > wet lab results (!!!) such cool work from Aadyot Bhatnagar et al biorxiv.org/content/10.110…

Asgard photos!!!!

Sometimes I think that proteins are truly motherfuckers like this one that binds Uranium ☢️💀 Protein with twin binding sites for uranium extraction from seawater academic.oup.com/nsr/advance-ar… img source synbiobeta.com/read/uranium-e…

x.com/TheRickyDavila… Recognize evil when it shows its face to you.

The ChimeraX 1.10 release candidate and daily builds can predict small complexes of proteins, nucleic acids and small molecules using Boltz on your Mac, Windows or Linux computer. rbvi.ucsf.edu/chimerax/data/…

We are thrilled to announce our new publication in Science Advances: Gaia, an AI-powered protein search platform that brings genomic context into functional annotation. science.org/doi/10.1126/sc… Gaia enables rapid, scalable discovery of remote homologs across 131,000+ genomes —