One of the key issues in structure-based drug discovery is prediction of the binding affinities of candidate ligand molecules to the target molecules. This is often referred to as the “scoring problem”. A whole spectrum of methods has been developed to solve this problem, and a group of approaches called “scoring functions” has gained popularity. A scoring function computes the fitness score of a ligand molecule to its target protein based on a given complex structure. These empirical scoring functions do not require extensive conformational sampling and are very fast in binding affinity prediction, and some of them were also found to have reasonable accuracy. For these reasons, they have extensively been applied in high-throughput virtual library screening and detailed molecular docking studies.
Researchers at the University of Michigan have developed X-Score, a computer program for computing protein-ligand binding affinities. The X-Score program computes the dissociation constant of a protein-ligand complex based on its three-dimensional structure. It uses an empirical equation that considers the important energetic factors in a protein-ligand binding process. This equation is calibrated with a number of protein-ligand complexes with known three-dimensional structures and binding affinities. When tested on an independent set of 30 protein-ligand complexes. X-Score is able to predict their binding free energies with a standard deviation of 2.2 kcal/mol.
Applications and Advantages
- Structure-based drug discovery: High-throughput virtual library screening and detailed molecular docking studies
- Improved docking accuracy as compared to the conventional force field computation used for molecular docking