Anna Vangone

MIMOSA improves de novo antibody design by constraining generative models to known binder interfaces, achieving validated binders on targets where unconstrained methods fail.

The paper develops a Random Forest classifier, CoDES, to distinguish correct from incorrect protein-protein docking models using machine learning and scoring functions.

The paper identifies mutations in the SARS-CoV-2 spike protein's fusion core that disrupt post-fusion assembly, with D936Y showing increased prevalence and weakening interactions.

This paper uses residue-residue distance predictions to assess protein flexibility, finding that flexible residue pairs have more local maxima in their predicted distance distributions.

The paper evaluates HADDOCK's performance in CAPRI rounds 38-45, showing it produces acceptable or better models for most targets using data-driven strategies.

iSEE predicts binding affinity changes upon mutations using a random forest model with 31 features, achieving high performance on diverse datasets.