Novel algorithm for efficient distribution of molecular docking calculations

Molecular docking is a computational method to study the formation of intermolecular complexes between two molecules. In drug discovery, it is employed to estimate the binding between a small ligand (the drug candidate), and a protein of known three-dimensional structure. Docking is becoming a standard part of workflow in drug discovery. Recently, we have used the software VINA, a de facto standard in molecular docking, to perform extensive docking analysis. Unfortunately, performing a successful blind docking procedure requires large computational resources that can be obtained by the use of clusters or dedicated grid. Here we present a new tool to distribute efficiently a molecular docking calculation onto a grid changing the distribution paradigm: we define portions on the protein surface, named hotspots, and the grid will perform a local docking for each region. Performance studies have been conducted via the software GRIMD.