PDB-CAT: A user-friendly tool to classify and analyze PDB protein-ligand complexes.

Area of Science:

• Structural Biology
• Bioinformatics
• Computational Chemistry

Background:

• The Protein Data Bank (PDB) is a critical resource for understanding protein-ligand interactions, but lacks automated tools for classifying structures.
• Differentiating between apo (ligand-free) and holo (ligand-bound) states, and identifying covalent versus non-covalent binding, is challenging.
• This hinders efficient retrieval and analysis of specific structural datasets for drug discovery and mechanistic studies.

Purpose of the Study:

• To develop PDB-CAT, a user-friendly tool for automated classification and information extraction from PDBx/mmCIF files.
• To accurately categorize protein structures based on ligand presence and binding type (apo, holo, covalent, non-covalent).
• To enable verification of protein sequence mutations against reference sequences.

Main Methods:

• PDB-CAT employs a parallelized implementation for efficient processing of PDBx/mmCIF files.
• It utilizes a blacklist-based approach to automatically identify ligands within protein complexes.
• Classification is based on ligand presence (apo vs. holo) and binding mode (covalent vs. non-covalent).

Main Results:

• PDB-CAT successfully classifies protein structures into apo, holo, covalent, and non-covalent categories.
• The tool can also identify mutations in protein sequences by comparison with reference data.
• Demonstrated efficiency by classifying SARS-CoV-2 Main Protease complexes and screening the PDBbindv2020 database in under 10 minutes.

Conclusions:

• PDB-CAT provides an efficient and automated solution for classifying protein structures in the PDB.
• The tool simplifies the retrieval of specific structural datasets, accelerating research in structural biology and drug discovery.
• Availability on GitHub and a tutorial on GitBook promote widespread adoption and use.