Knowledge-guided docking of flexible ligands to protein domains

Abstract

Study of protein interactions is important for the investigation of protein complex and for gaining insights into various biological processes. The conventional binding test in laboratory is very tedious and time-consuming. Therefore, computational methods are needed to predict possible protein interactions.

Protein docking is a computational problem that predicts possible binding between two molecules. Many algorithms have been developed to solve this problem. Rigid-body docking algorithms regard both molecules as rigid solid bodies and they are able to predict the correct binding efficiently. However, they are inadequate for handling conformational changes that occur during protein interactions. Flexible docking algorithms, on the other hand, regard molecules as flexible objects. Their performance is good when the size of the flexible molecule is relatively small. Larger flexible molecules increase the difficulty of the problem due to the large number of degrees of freedom.

A knowledge-guided flexible docking framework, BAMC, is presented. BAMC is targeted to protein domains with two or more well characterized binding sites that bind to relatively large ligands. There are three stages in BAMC: applying knowledge of binding sites, backbone alignment and Monte Carlo flexible docking. The first stage searches for binding sites of protein domains and binding motifs of ligands based on known features of the protein domain, and then constructs binding constraints. The second stage uses a backbone alignment method to search for the most favorable configuration of the backbone of the ligand that satisfies the binding constraints. The backbone-aligned ligands obtained serve as good starting points in the third stage which uses a Monte Carlo docking algorithm to perform the flexible docking.

BAMC has been successfully applied to three different protein domains: WW, SH2 and SH3 domains. Experimental results show that the BAMC framework is accurate and effective. The performance is better compared to AutoDock, a general docking program. Furthermore, using backbone-aligned ligands generated by BAMC as initial ligand conformations also improves the docking results of AutoDock.

BAMC has also been successfully applied to a benchmark set of 100 general test cases for protein-ligand docking. Experimental results show that the performance of BAMC is among the most consistent, compared to 9 existing protein docking programs. The performance of two docking programs is improved by using backbone-aligned ligands as input. Overall, the knowledge-guided approach adopted by the BAMC framework is important and useful in solving the difficult protein docking problem.