ATPbind is a meta-server approach to protein-ATP binding site prediction. Starting from given structure of query protein, ATPbind will identify the ATP-binding sites by using SVM to integrate the outputs of two template-based predictors, i.e., S-SITEatp (the extension of S-SITE) and TM-SITEatp (the extension of TM-SITE), and three discriminative sequence-driven features, i.e., position specific scoring matrix (PSSM), predicted secondary structure, and predicted solvent accessibility.
Users are also allowed to input primary sequence, where I-TASSER
will be used to generate 3D model first which are then fed into the ATPbind pipeline for protein-ATP binding site prediction. If the input is a sequence, we strongly suggest users to employ the I-TASSER server to model the 3D structure first, since the I-TASSER server used in ATPbind is not the newest one.
After protein-ATP binding site prediction, the ATPbind server implements a new binding pocket clustering scheme, PocHunter (download PocHunter.pdf
to see the details), to identify the pockets based on the predicted binding sites.
[View an example of output
[List the benchmark dataset
[Download the benchmark dataset
Jun Hu, Yang Li, Yang Zhang, Dongjun Yu.
ATPbind: accurate protein-ATP binding site prediction by combining sequence-profiling and structure-based comparisons.
Journal of Chemical Information and Modeling, 58(2): 501-510 (2018).
(Download PDF and