MT-ProtBERT: Multi-task Learning ProtBERT for Intrinsically Disordered Proteins Classification with Scarce Data
MT-ProtBERT consistently outperforms PARROT, an RNN-based IDP-specific model, across all tasks, and demonstrates that combining self-supervised and biochemistry-informed tasks, and multi-scale learning enables robust modeling of unstructured proteins under data scarcity.
Jian Sun, Kingshuk Ghosh, Lilianna Houston et al.
· 0 citations