Cross-Mapping of Protein - Ligand Binding Data Between ChEMBL and PDBbind

Area of Science:

• Biochemistry
• Structural Biology
• Cheminformatics

Background:

• ChEMBL is a comprehensive open data source for bioactive compound properties.
• PDBbind focuses on binding data for protein-ligand complexes from the Protein Data Bank.
• PDBbind's data volume is significantly smaller than ChEMBL's.

Purpose of the Study:

• To map and analyze the overlap between PDBbind and ChEMBL binding data.
• To determine the extent to which PDBbind data is a subset of ChEMBL.
• To understand the complementarity of focused versus comprehensive databases.

Main Methods:

• Mapping molecular information and binding data from PDBbind to ChEMBL records.
• Analyzing the intersection of binding data between the two databases.
• Investigating reasons for low data overlap.

Main Results:

• Only approximately 20% of PDBbind binding data has counterparts in ChEMBL.
• PDBbind data is largely complementary to ChEMBL data.
• Low overlap is due to limited ChEMBL coverage of PDBbind complexes and distinct literature screening.

Conclusions:

• PDBbind and ChEMBL databases offer largely complementary binding data.
• Focused databases like PDBbind provide unique value not fully captured by comprehensive databases.
• The study underscores the importance of both specialized and broad data resources in scientific research.