Validating Ligands with Phenix

TL;DR

This paper introduces a new tool in Phenix for validating ligands in protein structures using multiple metrics to improve drug discovery accuracy.

Contribution

A new tool in Phenix that combines multiple validation metrics for ligand modeling in crystallography.

Findings

The new tool evaluates ligand quality using model-to-data fit and geometric parameters.

It also considers stereochemistry, B-factors, hydrogen bonding, and steric clashes.

These metrics help identify modeling errors and increase confidence in ligand interpretation.

Abstract

Accurate modeling of ligands in macromolecular structures is essential for structure-based drug discovery. X-ray crystallography can provide such accurate models of protein-ligand complexes (1, 2). Since ligands typically contribute only a small portion of the total scattering signal and may bind with partial occupancy, their presence must be rigorously validated (3). Traditionally, ligand validation has relied on model-to-data fit metrics, such as real-space correlation to electron density maps in crystallography (Fig. 1). While such metrics remain important (4, 5), additional validation criteria can provide a more holistic assessment. This presentation will highlight a new tool in Phenix that evaluates ligand quality in crystallographic models using a combination of model-to-data fit, geometric parameters, stereochemistry, B-factors, hydrogen bonding, and steric clash analysis. We will demonstrate how these validation metrics can help identify modeling errors and increase confidence in ligand interpretation.