Eli Mann
@manntis4
Jesus follower, husband, atom simulator @RowanSci
Today, we’re releasing new data showing that Chai-2 can design antibodies against challenging targets with atomic precision. >86% of our designs possess industry-standard drug-quality properties without any optimization. Thread👇
Today we’re thrilled to announce JAM-2 — the first AI model capable of generating drug-quality antibodies straight from the computer, with industry-leading success rates. > Drug-like affinities: Picomolar to single-digit nanomolar antibody binders for half of 26 targets while…
Today, we're launching GPU-accelerated DFT for all Rowan users. Using GPU4PySCF, we see massive speedups for routine DFT tasks, making it significantly cheaper and faster to run high-accuracy quantum chemistry through Rowan. Here's two representative examples:
We're trying something new at Rowan—interviews! Bindwell just raised $6M to discover pesticides w/ AI. Today on the blog I ask Bindwell co-founder @navvye about pesticides, how AI gives them unfair advantages, and what new research he's excited about. rowansci.com/blog/a-convers…
Everyone is a computational chemist
Nike: Everyone is an athlete Apple: Everyone is an artist Shopify: Everyone is an entrepreneur Cursor: Everyone is a developer Cluly: Everyone cheats What's about your company?
We are not allowed to use MACE-OMOL owing to the license requirements! UMA does have issues with spurious negative frequencies, as do many other NNPs. @AriWagen did a study touching on this earlier: rowansci.com/blog/which-opt…
(5) And, finally, we're launching MSA generation as a workflow for Rowan subscribers today. If you're interested in securely generating MSAs w/o relying on public servers, consider using Rowan! We've written example scripts to show how easy this can be: rowansci.com/blog/using-row…
We're excited to launch a bunch of new features on Rowan today! (1) ion-mobility MS prediction (2) high-throughput docking (3) pose strain estimation (4) Lyrebird, a flow-matching conformer-generation model (5) & MSA generation for co-folding. Brief summaries to follow:
After two years of work, we’ve made an AI Scientist that runs for days and makes genuine discoveries. Working with external collaborators, we report seven externally validated discoveries across multiple fields. It is available right now for anyone to use. 1/5
Update: this might be the first Thanksgiving in which brussels sprouts don't even briefly overtake deep learning x.com/Miles_Brundage…
Deep learning will soon be more popular than Brussels sprouts.
BoltzGen, the new protein-binder-design model from @HannesStaerk and coworkers @MITCSAIL, is now live on Rowan for all users. BoltzGen is an open-source all-atom generative model for designing protein/peptide binders to a diverse array of targets—think Boltz-2, but generative.
New Rowan release! Today, we're launching: (1) pose-analysis MD for deeper analysis of protein–ligand binding modes following docking/co-folding . (2) a suite of ML-based microscopic pKa models in water and non-aqueous solvents. Read on for context, details, & screenshots.
In those days there was no king in Israel; everyone did what was right in his own eyes. — Judges 21:25
LLM engineered carbon capture enzymes have officially been produced. The best designs were 170% more active and 25% more stable across extreme pH (Tm +8.5 C). Winning strategies include adapting a tag from a bacterial carbonic anhydrase, beta barrel core packing, and removing…
We just posted a high-level overview of different strategies for in silico pKa prediction, with the goal of making this somewhat complex field a bit more accessible! Check it out here: rowansci.com/blog/how-to-pr…
on a quest to generate the most cursed structures ever to figure out what will and won't break co-folding for our users some of the structures are pretty funny—ever wonder what 4 ribonuclease inhibitors or 1500 alanines would look like?
Today, we added much-needed advanced search functionality to the @RowanSci platform! It's easy to run hundreds of jobs, but it used to be hard to find old things. You can now filter by name, status, type, starred status, visibility, created-at date, and last-updated-at date.
United States Tendencias
- 1. Comey 126K posts
- 2. GeForce Season 1,688 posts
- 3. Everton 79.2K posts
- 4. Dorgu 10.8K posts
- 5. Gueye 18K posts
- 6. Opus 4.5 5,454 posts
- 7. Mark Kelly 84.9K posts
- 8. Seton Hall N/A
- 9. Keane 12K posts
- 10. Halligan 39.9K posts
- 11. #MUNEVE 8,661 posts
- 12. Amorim 26.8K posts
- 13. Hegseth 30.3K posts
- 14. Department of War 24.8K posts
- 15. Pentagon 18K posts
- 16. #WooSoxWishList 12K posts
- 17. UCMJ 13K posts
- 18. Zirkzee 13.9K posts
- 19. Amad 7,964 posts
- 20. Creighton 1,995 posts
Something went wrong.
Something went wrong.