1/ At Valence Labs, @recursionpharma.bsky.social's AI research engine, we’re focused on advancing drug discovery outcomes through cutting-edge computational methods Today, we're excited to share our vision for building virtual cells, guided by the predict-explain-discover framework 🧵

(1/5) The Molecular Machine Learning Conference (MoML) is back @mila-quebec.bsky.social this June 18th. Join researchers advancing machine learning, molecular modeling, and therapeutic design. Tickets are free for students. Register Today: portal.ml4dd.com/events/molec...

Had a great time at the Ascona 2025 conference on benchmarking in biology! Thanks to @markrobinsonca.bsky.social and all other organizers. I’m grateful for the opportunity to present our @polarishub.io platform at the historic and beautiful Monte Verità (“Mount Truth”) venue. Look at that view! 😍

🚀 Antiviral Challenge Results! 🚀 After weeks of competition, and nearly 400 submissions, the antiviral challenge has concluded! Thank you to all of our participants. Each submission came with a report — check out the full leaderboard and find more details 👇

✨ Symposium session reveal ✨ AI-assisted Drug Discovery: Benchmarking the Field - Gustavo Stolovitzky and @caswognum.nl will explore how benchmarking can drive better, more reliable AI tools for drug discovery and biomedicine. Learn more and register here: conscience.ca/symposium2025

In case you missed it!

Polaris@polarishub.io · last yr.

Thanks to everyone who joined us last week! You can now watch the recording of the webinar: youtu.be/_ykP_tEVrOk

Join us tomorrow to hear @peterskrinjar.bsky.social talk about 🌹 Runs N' Poses 🌹! Curious to hear your thoughts on how to keep up with benchmarking in this field.

Polaris@polarishub.io · last yr.

We’re hosting @peterskrinjar.bsky.social for a webinar to discuss his latest paper “Have protein co-folding methods moved beyond memorization?”! Join live on Feb 21st at 11 AM ET. Register here: us06web.zoom.us/webinar/regi... Explore Runs N’ Poses on Polaris today: polarishub.io/benchmarks/p...

⚡ The intermediate leaderboard for the antiviral challenge is now live! 👀 See how you placed: antiviral.polarishub.io The challenge officially ends on March 10th, but there's one last leaderboard update before the final results. Submit before midnight on Feb 26th! polarishub.io/competitions

We’re hosting @peterskrinjar.bsky.social for a webinar to discuss his latest paper “Have protein co-folding methods moved beyond memorization?”! Join live on Feb 21st at 11 AM ET. Register here: us06web.zoom.us/webinar/regi... Explore Runs N’ Poses on Polaris today: polarishub.io/benchmarks/p...

Endless views and impressive cloudscapes! #photography #lions #kenya #maasaimara #simba

We’re extremely excited to host Runs N’ Poses on Polaris! The authors show how current co-folding methods struggle to generalize beyond ligand poses in their training data. Explore the benchmark: polarishub.io/benchmarks/p... Explore the dataset: polarishub.io/datasets/pli...

Peter Škrinjar@peterskrinjar.bsky.social · last yr.

Excited to share our latest preprint evaluating AlphaFold3, Boltz-1, Chai-1 and Protenix for predicting protein-ligand interactions, featuring our newly introduced benchmark dataset 🌹Runs N’ Poses🌹! www.biorxiv.org/content/10.1... 🧵👇 (1/n)

New leaderboard on @polarishub.io for Runs 'N Poses! 🎸 Anyone has any protein-ligand co-folding methods laying around they would like to put to the test? polarishub.io/benchmarks/p... Great work @peterskrinjar.bsky.social @jeeberhardt.bsky.social @torstenschwede.bsky.social @ninjani.bsky.social

Peter Škrinjar@peterskrinjar.bsky.social · last yr.

We’re pre-printing this early to get feedback from the community. We teamed up with @caswognum.nl at @polarishub.io to make the dataset and benchmark ML-ready polarishub.io/datasets/pli.... See also our github.com/plinder-org/... for more. Would love to hear your thoughts! (8/n

Excited to share our latest preprint evaluating AlphaFold3, Boltz-1, Chai-1 and Protenix for predicting protein-ligand interactions, featuring our newly introduced benchmark dataset 🌹Runs N’ Poses🌹! www.biorxiv.org/content/10.1... 🧵👇 (1/n)

Have protein-ligand co-folding methods moved beyond memorisation?

Deep learning has driven major breakthroughs in protein structure prediction, however the next critical advance is accurately predicting how proteins interact with other molecules, especially small mo...

biorxiv.org

Big milestone for @polarishub.io. We merged our first external contribution, thanks to @fastprop.bsky.social! 🎉 github.com/polaris-hub/... #opensource #benchmarking #ml #drugdiscovery

`Subset.as_dataframe()` hides targets instead of raising an error by JacksonBurns · Pull Request #258 · polaris-hub/polaris

Changelogs Subset.as_dataframe() would raise an error when called on a test set (which hides its targets), this PR allows method to work, just doesn't return targest Checklist: Was this PR...

github.com

🚨 Extension Alert: The intermediate leaderboard submission deadline for the antiviral challenge has been extended to midnight on February 7th! Don't miss this chance to see how you stack up against other participants! Submit your results today: polarishub.io/competitions

With @polarishub.io we early on decided to purposefully restrict the flexibility of our datasets and benchmarks. We understand that this can be frustrating at times, but we believe it’s an important step to improve reproducibility. Our team is here to help if you need support! Reach out on Github.

Polaris@polarishub.io · 2y ago

You’ve heard us talk about our mission to be the source of truth for the ML community in drug discovery. But what does that really mean, and how does it shape the way Polaris is built? In the latest blog, we break down the key design choices that make this possible: polarishub.io/blog/reprodu...

💡Have questions about the data behind the antiviral challenge?💡 This week, @asapdiscovery.bsky.social and @omsf.io are hosting 3 virtual office hour sessions with the experimentalists behind the data! RSVP here: docs.google.com/forms/d/e/1F...

Only a few days left to submit your predictions for the first, intermediate leaderboard. We’ve got an incredibly passionate team on standby to help solve any issues you may run into and we’d love to hear your feedback. 🧑‍💻 Github: github.com/polaris-hub/... 💬 Discord: discord.gg/vBFd8p6H7u

Polaris@polarishub.io · 2y ago

📣 We heard your feedback! We’ll be putting all results submitted by Jan 31st on an intermediate leaderboard so you can see how you rank across other participants in the antiviral challenge. Ready to put your predictions to the test? 📨 Submit your results today: polarishub.io/competitions

If you have been waiting for the right time to start working on the ASAP Discovery challenge, this is it! 🏆 For all results submitted up until January 31st, we will publish an intermediate leaderboard. Get your first set of results in now and see how you stack up against your peers.

Polaris@polarishub.io · 2y ago

📣 We heard your feedback! We’ll be putting all results submitted by Jan 31st on an intermediate leaderboard so you can see how you rank across other participants in the antiviral challenge. Ready to put your predictions to the test? 📨 Submit your results today: polarishub.io/competitions

The @polarishub.io discord is feeling very lively right now. I love learning with and from all of you!💡 If you haven't yet, join us and put your methods to the test to learn what is working and what's not in ADMET, potency and pose prediction. Get started in a few lines of code. Great video by Lu!

Polaris@polarishub.io · 2y ago

It only takes a few lines of code to get started with the @asapdiscovery.bsky.social x @omsf.io antiviral challenge! In this short tutorial, we’ll show you how easy it is to login, load the challenge data, train a simple baseline model, and submit your predictions 🧵 youtu.be/KBZJFA_arAg

Incredibly excited to see how well ADMET, affinity, and pose prediction models *actually* perform in the ASAP x OpenADMET x PolarisHub blind challenge: polarishub.io/competitions

Polaris

The benchmarking platform for drug discovery

polarishub.io

Working on predictive models for drug discovery? Here's your chance to see how well your models actually work! Together with @polarishub.io and OpenADMET, we've launched a blind prospective challenge to see how well predictive models for drug discovery can actually predict real drug discovery data!

Polaris@polarishub.io · 2y ago

🏁 The antiviral challenge is live! 🏁 Ready to test your skills on new data? Hosted in partnership with @asapdiscovery.bsky.social and @omsf.io, we've prepared detailed notebooks showcasing how to format your data and submit your solutions. 🧑‍💻

The @asapdiscovery.bsky.social x OpenADMET (@omsf.io) blind challenge is live! 🚀 You're free to use any publicly available training data, but the challenge provides 770 ligand poses, 1031 potency datapoints and 434 ADMET datapoints as a consistently generated training set. Come learn with us! 🏆

Polaris@polarishub.io · 2y ago

🏁 The antiviral challenge is live! 🏁 Ready to test your skills on new data? Hosted in partnership with @asapdiscovery.bsky.social and @omsf.io, we've prepared detailed notebooks showcasing how to format your data and submit your solutions. 🧑‍💻

🎥 At NeurIPS, we had the opportunity to chat with the authors of several dataset and benchmarking papers that are relevant to machine learning for drug discovery. First up—a quick interview with the team behind WelQrate! 🧵 youtu.be/pbatKLphJAE

NeurIPS 2024 Series - A Chat with the WelQrate Team

YouTube video by PolarisHQ

youtu.be

🎉 Zarr-Python 3 is here! 🎉 - Full support for Zarr v3 spec - Chunk-sharding for more efficient data storage - Major performance boosts with async I/O & parallel compression 💻 pip install --upgrade zarr 💻 conda install --channel conda-forge zarr Blog post: https://buff.ly/3C3OwYw

Hi, Zarr community! 👋 🚨 Big news: Zarr-Python 3.0 is dropping next week! 🚨 If your project depends on Zarr-Python, make sure to check out the migration guide to get ready for the update. Guide: https://buff.ly/3W0cLxd More updates soon! 🧑🏻‍💻

Vaderland! 🇳🇱 #netherlands #photography

Scar! 🦁 #lion #photography #kenya #maasai #mara

To change benchmarking practices in ML for drug discovery, I've always felt strongly we need to go beyond publishing papers. Excited to share a first glimpse into what that could look like! 👀

Polaris@polarishub.io · 2y ago

📊 Imagining the Future of ML Evaluation in Drug Discovery Our recent paper discussed the limitations of static leaderboards—they never tell the full story. What if we had a better and easier way of evaluating methods? A vision for the future, in the latest blog 🧵 polarishub.io/blog/imagini...

This is exactly our strategy with @polarishub.io! We want to provide ML-ready datasets that make it easy to do the right thing and hard to do the wrong thing.

David Ryan Koes@dkoes.compstruct.org · 2y ago

Agree with having all the data available, don’t agree with requiring extra steps to get “good” training set - should require extra steps to get all data (eg like you do with columns). I talked with too many ML researchers who were using molecular data uncritically this week: make default reasonable.