Contrastive learning for antibody-antigen sequence-to-specificity prediction
Hyunho Lee, Karla M. Castro, Sean Renwick, Lucas Stalder, Wiona Sophie Glänzer, Rachita K. Kumar, NingNing Chen, Andreas Scheck, Alexander Yermanos, Derek M. Mason, Sai T. Reddy
This paper introduces CALM, a model that predicts antibody-antigen binding from sequences using contrastive learning, achieving 7% top-1 retrieval on a test set.
bioRxiv · Feb 26, 2026 · PreprintImmunology