Whole-genome prediction of bacterial pathogenic capacity on novel bacteria using protein language models, with PathogenFinder2
Alfred Ferrer Florensa, José Juan Almagro Armenteros, Rolf Sommer Kaas, Philip T. L. C. Clausen, Henrik Vedel Nielsen, Burkhard Rost, Frank M. Aarestrup
A ProtT5-embedding, ConvNext-plus-attention ensemble predicts human pathogenic capacity from whole bacterial genomes, outperforming prior predictors on species absent from training and flagging influential proteins.
Bioinformatics · Mar 21, 2026Peer-reviewedBioinformatics✦ In-depth digest