Usage of this web interface: Two PDB files whose ligands need to be compared are obtained from the upload. Each of the protein files is then processed to extract all ligands and their corresponding site files. All vs All comparison of sites is done with PocketMatch and scores are reported. If there are M ligands/sites in one protein and N in another, then number comparisons is M*N. Grouping of residues is an important factor in computing the score (see References section for more details)

Upload two PDB files whose pockets need to be compared: Your uploads should match the PDB file format as closely as possible. All lines should contain 80 characters. Your input files should contain ATOM protein records only. All ATOM records for nucleotides should be removed. HETATM and ANISOU records can be processed, but are discouraged. Example pockets are given below:

example pocket #1

example pocket #2

Upload pocket #1 (PDB format)  

Upload pocket #2 (PDB format)  

Database vs. database comparisons are possible with a local installable version of PocketMatch. Executables/binaries can be downloaded from below:

PocketMatch version 1.0

PocketMatch version 1.1

PocketMatch version 2.0

PocketMatch version 2.1


Reference #1: Yeturu, Kalidas, and Nagasuma Chandra. "PocketMatch: a new algorithm to compare binding sites in protein structures." BMC bioinformatics 9.1 (2008): 543.

Reference #2: Nagarajan, D., & Chandra, N. (2013, February). PocketMatch (version 2.0): A parallel algorithm for the detection of structural similarities between protein ligand binding-sites. In Parallel Computing Technologies (PARCOMPTECH), 2013 National Conference on (pp. 1-6). IEEE.