PDB-Explorer: a web-based interactive map of the protein data bank in shape space

Area of Science:

• Structural Biology
• Bioinformatics
• Computational Chemistry

Background:

• The Protein Data Bank (PDB) offers access to experimentally determined 3D structures of biological macromolecules.
• Current tools for exploring the PDB have limitations in accessing global structural diversity and identifying relationships between structures.

Purpose of the Study:

• To develop a novel method for representing and comparing the 3D shapes of PDB entries.
• To create an interactive tool for visualizing and exploring the structural diversity of the entire PDB.

Main Methods:

• A 136-dimensional atom pair 3D-fingerprint (3DP) was designed to capture protein molecular shape.
• Nearest neighbor searches and principal component analysis were used to analyze and visualize PDB structures in 3DP-space.
• An interactive website, PDB-Explorer, was developed using JavaScript for platform independence.

Main Results:

• The 3DP fingerprint effectively groups proteins by 3D shape similarity while distinguishing closely related structures.
• PDB-Explorer provides an interactive map of PDB structures, enabling visualization and nearest neighbor searches.
• The tool offers instantaneous similarity searches comparable to slower 3D alignment methods and clusters proteins by superfamily.

Conclusions:

• A novel chemical space classification of the PDB based on molecular shape has been achieved using the 3DP fingerprint.
• The PDB-Explorer website provides an unprecedented, interactive platform for exploring the structural diversity of the PDB.
• The tool is freely available and facilitates the visualization and analysis of biological macromolecular structures.