The COACH-D modeling results are generally summarized in a webpage, the link of which is sent to the users after the modeling is completed (see an example of the COACH-D output). This page includes a detailed explanation on the data listed on the COACH-D output page.



The input to COACH-D can be either the sequence or structure of the query protein. Users are also allowed to submit query ligand for docking. As shown in Figure 1, the COACH-D works as follows.
(1) When the sequence of query receptor is given, the I-TASSER Suite will be used to predict the structure for the receptor.
(2) The predicted structure is submitted to five individual methods (TM-SITE, S-SITE, COFACTOR, FINDSITE and ConCavity ) to predict the protein-ligand binding pockets and residues. Among them, four make predictions by searching the query structure/sequence through the BioLiP template library.
(3) The template ligand and the ligand submitted by the users are docked into the predicted binding pockets, using the efficient docking algorithm AutoDock Vina.

Figure 1. The flowchart of the COACH-D algorithm.

Submit a job to COACH-D

Figure 2. The submission page of COACH-D.

Download the complete set of results

Figure 3. The tarball of the COACH-D results.

The predicted ligand-binding sites and poses

Figure 4. The tarball of the COACH-D results.

How to cite COACH-D?

Please cite the following article when you use the COACH-D server:
  • Q Wu, Z Peng, Y Zhang, J Yang, COACH-D: improved protein-ligand binding site prediction with refined ligand-binding poses through molecular docking, Nucleic Acids Research, 46: W438–W442 (2018).
  • J Yang, A Roy, Y Zhang, Protein-ligand binding site recognition using complementary binding-specific substructure comparison and sequence profile alignment, Bioinformatics, 29: 2588-2595 (2013).
  • Need more help?

    If you have more questions or comments about the server, please email