Exploring cooperative molecular contacts using a PostgreSQL database system

Area of Science:

• Computational Biology
• Structural Bioinformatics
• Biochemistry

Background:

• Cooperative molecular contacts are crucial for protein structure and function, particularly in ligand binding.
• Understanding these interactions requires efficient methods for analyzing atomic environments and relationships within proteins.

Purpose of the Study:

• To create a flexible PostgreSQL database for mining molecular contacts based on atomic environments.
• To investigate the role of carboxylate groups in the Ser-His-Asp/Glu catalytic triad and identify potential catalytic functions in unannotated proteins.

Main Methods:

• Construction of a PostgreSQL database storing protein structural information as atomic environments.
• Flexible querying of the database to mine molecular contacts, focusing on catalytic triads (Ser-His-Asp/Glu).
• Inclusion of ligand atoms (using Sybyl atom types) for analyzing interactions, exemplified by evaluating counter-ion proportions for ligand carboxylates.

Main Results:

• The presence of carboxylate oxygen atoms near histidine residues was associated with shorter Ser-OH..N-His hydrogen bonds in the PDB30 dataset.
• Prospective mining of catalytic triads in unannotated proteins suggested potential catalytic functions for these proteins.
• The database system successfully incorporated ligand atom information, enabling analysis of ligand-protein interactions like counter-ion binding.

Conclusions:

• The developed database provides a powerful tool for flexible mining of molecular contacts and protein structural analysis.
• The findings highlight the influence of carboxylate proximity on catalytic triad geometry and suggest a method for functional annotation of proteins.
• The system's ability to include ligand atoms broadens its applicability to studying drug-target interactions and other ligand-mediated processes.