BioLiP is a semi-manually curated database for high-quality, biologically relevant ligand-protein binding interactions. The structure data are collected primarily from the Protein Data Bank, with biological insights mined from literature and other specific databases. BioLiP aims to construct the most comprehensive and accurate database for serving the needs of ligand-protein docking, virtual ligand screening and protein function annotation. Questions about the BioLiP Database can be posted at the Service System Discussion Board.

Since ligand molecules (e.g., Glycerol, Ethylene glycol) are often used as additives (i.e., false positives) for solving the protein structures, not all ligands present in the PDB database are biologically relevant. BioLiP uses a composite automated and manual procedure for examining the biological relevance of ligands in the PDB database. Each entry in BioLiP contains a comprehensive list of annotations on:

ligand-binding residues

ligand binding affinity (from the original literature, plus Binding MOAD, PDBbind-CN, BindingDB),

catalytic site residues (mapped from Catalytic Site Atlas)

Enzyme Commission numbers

Gene Ontology terms

cross-links to the PDB, UniProt, PDBsum, PDBe, and PubMed databases.

To annotate the function of uncharacterized proteins using the BioLiP database, we developed a new algorithm COACH to predict ligand-binding sites. The COACH server is available at https://zhanglab.dcmb.med.umich.edu/COACH/

BioLiP in numbers

BioLiP is updated weekly (every Friday) and the current version (Apr 01, 2022) contains: Number of entries: 573,225 Number of proteins from PDB: 116,643 Number of DNA/RNA ligands: 60,539 Number of peptide ligands: 28,086 Number of metal ligands: 156,972 Number of regular ligands: 327,620 Number of entries with binding affinity data: 23,492 (10,971 from Binding MOAD, 16,980 from PDBbind-CN, 7,331 from BindingDB, and 64 from manual survey of the original literature) The ocurence frequency of all ligands in BioLiP.

BioLiP news

Current version: Apr 01, 2022 with 573,225 entries. 2014-09-10: The ocurence frequency of all ligands in BioLiP was provided on the ligand information page. 2014-09-10: A Perl script (download_all_sets.pl) was provided for automatic download of all the weekly updated subsets. 2013-08-19: A link was added at the search and browse pages to allow users to export searching results in CSV format. 2013-03-15: BioLiP was rebuilt with the following changes: (1) All DNA/RNA-protein interactions from PDB were added, which should be useful for studying Nucleic Acids-protein interactions. In previous versions, only small DNA/RNA ligands were considered. (2) The k-mer ligand "UUU" was divided into component ligands and removed from the database. 2012-12-14: BioLiP was updated with 211,382 entries. 102 ligands were added to the artifact ligand list, which helped us reduce about 1000 false positive entries. 2012-12-07: BioLiP was updated with 211,819 entries. 642 false postive entries for the ligands RG1 and IMD were manually removed. 2012-11-30: BioLiP was updated with 212,148 entries. The COACH algorithm was redesigned and is available at here. 2012-11-16: BioLiP was updated with 211,040 entries. 233 false postive entries for the ligands FTT and MYR were manually removed. 2012-11-09: BioLiP was updated with 210,894 entries. The number of BioLiP entries with binding affinity data was increased from 20127 to 22091 with the latest realse of the PDBbind-CN database. 2012-11-02: BioLiP was updated with 210,377 entries. 108 false postive entries for the ligands TRD and UND were manually removed. >> Read more about BioLiP news...