Fig. 3
From: Sequence-based prediction of protein binding regions and drug–target interactions
Prediction and visualization of binding regions on homologous 3D complexes with low sequence similarity. We visualized BR prediction for beta-lactamase in the test dataset (PDB ID: 1GHM, UniProt accession: P00807, white colored), which is homologous to the training protein (PDB ID: 5EEC, UniProt accession: Q9F663, dark gray colored) with low sequential identity/similarity. a Structural alignment results for 5EEC and 1GHM. True BRs of 5EEC are colored red, and the predicted top-3 BRs are colored blue for IGHM. b–d Structural and sequential alignment of predicted BRs, showing that HoTS can predict the same BRs with different residues