Avoiding pitfalls when modelling ligands in macromolecular crystallography

Area of Science:

• Structural Biology
• Drug Discovery
• Biochemistry

Background:

• X-ray crystallography is crucial for determining protein-ligand complex structures in drug discovery.
• Flawed ligand models in the Protein Data Bank (PDB) can lead to misdirected research efforts.

Purpose of the Study:

• To outline a practical approach for validating ligand structures during X-ray crystallography.
• To highlight common pitfalls in ligand modeling and deposition.

Main Methods:

• Application of validation tools including Coot, MolProbity, Mogul, and Buster-report.
• Re-examination of deposited PDB entries to identify modeling errors.
• Emphasis on a 'null-hypothesis' approach and continuous critical assessment.

Main Results:

• Identified key pitfalls: modeling into ambiguous electron density, incorrect chemical definitions (chirality, tautomers), poor ligand fit, and data/model mismatches.
• Demonstrated the utility of validation tools in preventing these errors.

Conclusions:

• Rigorous ligand validation is essential for improving the reliability of deposited protein-ligand complex structures.
• Continuous critical assessment throughout the modeling process is paramount for accurate structure-guided drug discovery.