Cédric Bouysset

User Webinar: Improving Pose Prediction Models for Drug Discovery. Join us for an insightful webinar with David Errington and Cedric Bouysset from Recursion, the authors of a recent study that challenges the current benchmarks in protein-ligand pose prediction.

David Errington and Cédric Bouysset @recursionpharma.bsky.social recently published a study that challenges the current benchmarks in protein–ligand pose prediction.

Register for the upcoming webinar to learn more: attendee.gotowebinar.com/register/599...

⌛9th Sept - 15:00 (BST)

#DrugDiscovery

"Celebrities turning 40 this week"

We believe that incorporating interaction-aware metrics into evaluation pipelines is essential for advancing the field, particularly for drug discovery applications. This could involve developing PLIF or pharmacophore-sensitive loss functions for training new ML models.

- While cofolding models show promise in predicting the full complex structure, our findings suggest a significant need for improvement in ensuring the predicted poses form meaningful interactions with the target protein.

- Our study shows that classical docking algorithms, when benchmarked with proper structure preparation, generally outperform current machine learning-based methods in recovering interactions.

- We introduce the Protein-Ligand Interaction Fingerprint (PLIF) recovery as an additional metric to directly assess the recall of these vital interactions.

Key takeaways:
- Focusing solely on ligand placement accuracy can overestimate model performance. We found that many predicted poses, especially from recent protein-ligand cofolding models, fail to capture key interactions.

We explore a critical aspect often overlooked in recent protein-ligand pose prediction benchmarks: the accurate recapitulation of protein-ligand interactions. While low RMSD and physical plausibility are important, they don’t tell the whole story.

Our work “Assessing interaction recovery of predicted protein-ligand poses” with David Errington, @cschneider.bsky.social and @fdreyer.bsky.social, previously presented at NeurIPS, is now published in the Journal of Cheminformatics! 🎉
Check out the full paper here: rdcu.be/emKgd

MDAnalysis UGM logo

#mdaUGM2025 is officially happening! Abstracts and travel bursary applications are being accepted from now until the July 15, 2025 deadline: www.mdanalysis.org/2025/04/13/u....

🗓️ November 9-11, 2025
📍Tempe, Arizona, USA (and online)

#open-source-software #molecular #simulations #streaming

Photo of the inside of a university building at WUR, it looks like a mix of a greenhouse with offices that have balconies.

We're offering a fully funded PhD at the intersection of ML/AI and the natural sciences with a focus on sustainability and chemistry.

You'll work at WUR in the Netherlands, ranked #3 in environ. sciences, #1 in agricultural science, #38 in life sciences (QS).

Apply here:
www.wur.nl/nl/vacature/...

Happy to share that I'll be once again mentoring Google Summer of Code applicants with MDAnalysis 🤠 This year, alongside pure-MDAnalysis projects, we have collaborative projects on MolecularNodes, WESTPA... and 3 projects ideas for ProLIF ✨ More info in the post below 👇

PhD student - Machine learning in bio- and cheminformatics

Here's a first:
I'm hiring a PhD student.
If you have (or know someone who has) a strong background and interest in computational (bio)chemistry or computer science (especially 3D computer vision), let's work together 😊

www.wur.nl/nl/vacature/...

var(){ _var=$(echo "$@" | perl -pe 's/(.+?) *= *(.+)/\1=\2/') && declare -g $_var;}

It's annoying that the `declare` command doesn't work like `set` in csh which allows `set var = value`, but with a little bit of dubious trickery:

"He's a little confused, but he's got the spirit"