Dina Schneidman

11/ Finally, we will present a RaCoon poster at the Computational structural biology EMBO workshop this week: www.embl.org/about/info/c...

10/ RaCoon is fully open-source with a public web server.
Try it → evaluate variant pathogenicity with interpretable, residue-aware probabilities.
bio3d.cs.huji.ac.il/webserver/ra...

9/ Why does this matter?
Mis-calibrated probabilities = inconsistent evidence for clinical classification (and you may also improve AUROC).

8/ 🎯 Performance
RaCoon improves ESM1b substantially:
• ClinVar AUROC: 0.930 → 0.941
• ProteinGym AUROC: 0.912 → 0.924
• Per-protein AUROC improves too
• Calibration error (ECE/MCE) drops across all subgroups

7/ RaCoon 4-step pipeline:

1. Split variants by key residue properties
2. Fit GMMs to benign vs. pathogenic scores in each subgroup
3. Convert raw ESM1b LLRs into calibrated probabilities
4. No direct label exposure → labels only used for prior estimation

6/ RaCoon performs multicalibration of ESM1b, producing reliable probabilities across all relevant subgroups using minimal supervision.

racoon

5/ This motivated RaCoon (Residue-aware calibration via conditional distributions).
(It calibrates like a raccoon sorts trash: by categories. 😉)

4/ Surprisingly calibrating per variant subgroup (i.e. interface) not only improves miscalibration but also increase global AUROC across most models.

3/ We calibrate VEPs at the residue level since variant effects depend strongly on local residue properties. We find that model entropy distribution can guide calibration.

ECE

2/ In a calibrated model a model score of 0.8 means ~80% of similar variants are pathogenic. But VEPs often fail, especially for variants in disordered regions or protein-protein interfaces.

Calibrated Variant Effect Prediction at the Residue Level Using Conditional Score Distributions Effective clinical use of variant effect prediction (VEP) requires models that are both accurate and well-calibrated. Calibration refers to a model’s ability to produce meaningful and reliable probabi...

1/ Missense variant effect predictors (VEPs) are very accurate but are they trustworthy?
Not if their probability outputs aren’t calibrated.
Our new method, RaCoon, fixes this by calibrating VEPs at the residue level, where miscalibration actually happens. www.biorxiv.org/content/10.1...

Announcing CAPRI round 61 with four antibody-antigen targets - registration is open and the prediction period starts on Monday September 29th. Put your methods to the test! Visit… | Alexandre Bonvin Announcing CAPRI round 61 with four antibody-antigen targets - registration is open and the prediction period starts on Monday September 29th. Put your methods to the test! Visit https://lnkd.in/ei68T...

Still time to register for the next CAPRI round and put your methods to the test for predicting the structure of four antibody-antigen complexes - prediction period opening on Sept. 29th.

The second annual ALADDIN project meeting has come to an end! 👏

We thank our hosts from @liu.se for a well-organised meeting and our project partners for their presentations & insightful conversations.

As we approach our halfway milestone, things will only speed up!

#EICPathfinder

Unfortunately it’s clear that the Israeli people can’t topple the government or change its ways. Seems like the attorney general, the last gate keeper, will be fired soon. As long as we’re stuck with this government, the war won’t stop. Only international intervention (=the U.S) might do it.

CAPRI Docking

A small reminder to all structural biologists around working on biomolecular complexes: please consider sharing your complexes as targets for CAPRI - AI has not solved all structure prediction problems and there are still challenges! See www.capri-docking.org/contribute/

Atlas of imprinted and allele-specific DNA methylation in the human body - Nature Communications DNA methylation drives parental expression differences at imprinted genes. Here, the authors uncover hundreds of ubiquitous and tissue-specific differentially methylated regions, offering insights int...

The first comprehensive atlas of allele-specific DNA methylation
www.nature.com/articles/s41...