The scientific python ecosystem is amazing nowadays! 🎉Here is a PyRosetta interactive helix bundle editor written in 70 lines of code running on google colab and using ipywidgets and py3dmol. Try it here: colab.research.google.com/drive/1K30OqWo…

Gradio blocks and 3Dmol.js power the interactive visualization. I built the interface in a few hours (mostly tinkering with 3D vis) and could directly couple it with the models. Really nice and quick to setup with explicit dependencies so it won't break like Google Colab. 3/4

Very excited for tomorrow's paper! Fast and accurate conformer generation is now possible thanks to diffusion models on torsion angles! Join us at 11AM EST to discuss with the authors hannes-stark.com/logag-reading-…

If you missed it, the talks and slides are now ONLINE!! ebi.ac.uk/training/event…

Protein structure can be predicted from a single sequence alone with high accuracy. Helixon Bio team have developed OmegaFold, achieving performance similar to RF and AF2's MSA versions. Only a single sequence is given as input. 1/5

a really simple and powerful paper by @Schrodinger. TLDR: AlphaFold2 produces proteins at quality enough to do early enrichment comparable to apo/holo forms.

Thanks to Andrew McNutt (David Koes group), who converted #GNINA’s models’ weights from #Caffe to #PyTorch, 🔥 gnina-torch 🔥 (a PyTorch implementation of GNINA’s scoring function) now ships with all pre-trained GNINA’s models! 🧵 1/6 github.com/RMeli/gnina-to…

We are excited to announce that our inaugural 2022 Molecular ML conference site is now live! 😃 Registrations & poster submissions are now open: moml22.jclinic.mit.edu (Students register for free! 🌟 See More Info+ @ registration)

We've found oce to be immensely helpful for our custom software, products (Oloren Analysis Suite), and modeling efforts. We aim to bring together other public codebases in a cohesive unified manner, accelerating the adoption of AI in drug discovery. github.com/Oloren-AI/olor…

🔥 gnina-torch 🔥 is now available Hugging Face. Just drag-and-drop your protein-ligand complex and score it with #GNINA pre-trained models! Try it out: huggingface.co/spaces/RMeli/g… Powered by Hugging Face Gradio PyTorch pytorch-ignite 3Dmol.js (David Koes) 🧵 1/4

A new paper just published in JCIM & JCTC Journals! Auto3D: Automatic Generation of the Low-Energy 3D Structures with ANI Neural Network Potentials. Collaboration with Adrian Roitberg 🏳️‍⚧️ #compchem #MachineLearning pubs.acs.org/doi/full/10.10…

Excited to share DiffDock, new non-Euclidean diffusion for molecular docking! In PDBBind, standard benchmark, DiffDock outperforms by a huge margin (38% vs 23%) the previous state-of-the-art methods that were based on expensive search! arxiv.org/abs/2210.01776 A thread! 👇

Excited to share BigBind: Learning from Nonstructural Data for Structure-Based Virtual Screening doi.org/10.26434/chemr…. With Paul Francoeur, Rishal Aggarwal, Konstantin Popov, David Koes, and Alex Tropsha

🚨🚨🚨 We're looking for our next Chair! Please share this exciting opportunity to lead our cutting-edge Comp & Sys Bio Dept. at Pitt School of Med cfopitt.taleo.net/careersection/…

RT @GabriellaGerla1: TECBio REU @ Pitt Applications are open now!! This is a great program to do computational biology research during undergrad, and determine if you want to pursue a PhD. tecbioreu.pitt.edu pic.x.com/2qNwL26Vyw

We're hiring a machine learning engineer at the OpenFold Consortium to help manage the OpenFold codebase and ecosystem. Someone with a knack for engineering, science, and outreach. Competitive compensation. Look here if interested! (first job description): omsf.io/about/careers/

I’ll be sharing my recent work on all-atom protein structure prediction with SidechainNet at tonight’s COMP poster session 7-9pm at ACS. If you’re attending, come by! 10 (!!) posters from our research group will be on display covering deep learning and drug discovery. #ACS2023

As a part of the #llmhack, me and Sean Warren developed MolVerse, a language interface to 3dmol.js: a web app for all chemists and biologists that want to visualise their (bio)molecules without needing to code. Here we load the structure of benzene:

Excited to share PLANTAIN: arxiv.org/abs/2307.12090. Our docking approach applies diffusion-inspired training to the classic idea of pose score minimization. We achieve state-of-the-art results while being 10x faster than others. W/ Konstantin Popov, David Koes, and Alex Tropsha

bsky.app/profile/dkoes.…