Preprint—not peer reviewed
Structure-free, site-resolved contrastive learning extends small-molecule discovery beyond the reach of structure-based modeling
Ptarmigan-1 co-embeds protein residues and candidate small molecules from sequence and two-dimensional chemistry without constructing a binding pose. The authors report that it localizes reversible and covalent inhibitors on withheld targets and screens the human proteome against a compound library in under a day. A pose-free index could extend virtual screening to cryptic and disordered sites where well-defined pockets are unavailable.
AI–biochemistryAI–chemistry Molecular representation