Undergraduate Wang Jue has published an article titled “Protein-small molecule binding site prediction based on a pre-trained protein language model with contrastive learning” in the Journal of Cheminformatics. This study presents a protein-small molecule binding site prediction model (CLAPE-SMB) based on a pre-trained protein language model and contrastive learning, and organizes the protein-small molecule binding site dataset UniProtSMB. CLAPE-SMB achieved an MCC of 0.699 on the UniProtSMB test set, outperforming other models. Compared to structure-based prediction methods, CLAPE-SMB is particularly suitable for proteins that lack precise experimental structures, such as intrinsically disordered proteins (IDPs). This provides new insights into protein-small molecule interactions and offers new possibilities for drug design.
For detail: https://doi.org/10.1186/s13321-024-00920-2