iMOLSDOCK : induced-fit docking using mutually orthogonal Latin squares (MOLS)

TL;DR

iMOLSDOCK is a novel flexible receptor docking method that employs mutually orthogonal Latin squares to efficiently sample conformations, outperforming existing tools in accuracy despite higher computational cost.

Contribution

This paper introduces iMOLSDOCK, an extension of MOLSDOCK, incorporating flexible receptor modeling using MOLS and mean field analysis, which improves docking accuracy.

Findings

iMOLSDOCK outperforms GOLD and AutoDock Vina in accuracy.

The method effectively samples ligand and receptor conformations.

It requires more computational time than existing algorithms.

Abstract

We have earlier reported the MOLSDOCK technique to perform rigid receptor/flexible ligand docking. The method uses the MOLS method, developed in our laboratory. In this paper we report iMOLSDOCK, the 'flexible receptor' extension we have carried out to the algorithm MOLSDOCK. iMOLSDOCK uses mutually orthogonal Latin squares (MOLS) to sample the conformation and the docking pose of the ligand and also the flexible residues of the receptor protein. The method then uses a variant of the mean field technique to analyze the sample to arrive at the optimum. We have benchmarked and validated iMOLSDOCK with a dataset of 44 peptide-protein complexes with peptides. We have also compared iMOLSDOCK with other flexible receptor docking tools GOLD v5.2.1 and AutoDock Vina. The results obtained show that the method works better than these two algorithms, though it consumes more computer time.