For your convenience, you can use this Perl script for automatic download of all the datasets:
download_all_sets.pl.
Non-redundant dataset is a subset of the redundant dataset by protein sequence clustering at 90% sequence identity.
For both redundant and non-redundant datasets, the following files are provided:
Receptor: 3D structure of proteins interacting with at least one biologically relevent ligand; original PDB residue numbering.
Due to hard disk space constraints, BioLiP no longer provides separate batch download for receptor structures with residue number starting from 1. However, we provide a C++ program at receptor1.cpp to facilitate the generation of such files by the user.
Ligand: 3D structures of biologically relevant ligands