Preparing a database of corrected protein structures important in cell signaling pathways.

Area of Science:

• Structural Biology
• Computational Chemistry
• Drug Discovery

Background:

• Accurate macromolecular structures are crucial for structure-based drug design.
• X-ray crystallography can yield low-resolution structures, hindering atom differentiation (e.g., NH vs. O) and leading to missing amino acids.
• Existing protein structure data often requires refinement for reliable use in drug design protocols.

Purpose of the Study:

• To develop a curated database of corrected 3D protein structure files.
• To enhance the quality and completeness of protein structures frequently utilized in structure-based drug design.
• To provide researchers with reliable structural data for improved drug discovery efforts.

Main Methods:

• Collected 3454 soluble cancer signaling proteins from the Protein Data Bank (PDB).
• Selected a subset of 1001 proteins for structural refinement, including bond order adjustment, formal charge correction, and addition of missing side chains.
• Employed homology modeling to correct missing backbone residues in 12 proteins and utilized molecular dynamics simulations (30 ns) to validate model stability.

Main Results:

• Successfully corrected 896 out of 1001 protein structures.
• Homology modeling yielded acceptable models for 12 proteins with missing residues, validated by Ramachandran, z-score, and DOPE energy plots.
• Molecular dynamics simulations confirmed the stability of the refined protein models (RMSD, RMSF, Rg values).

Conclusions:

• A refined database of 1001 protein structures with corrected atomic details and missing residues was generated.
• The database, including homology-modeled and molecular dynamics-simulated structures, offers improved reliability for drug design.
• The database will be expanded with more water-soluble proteins and made publicly accessible online.