Section: New Results

Critical assessment of protein-ligand docking methods via Drug Design Data Resource (D3R) Challenge 2015

Participants : Andreas Eisenbarth, Sergei Grudinin.

We participated at the Drug Design Data Resource (D3R) Challenge 2015. In the challenge, we were given protein structures and sets of ligand molecules in order to detect the putative binding poses. The aim was to find the energetically most favourable pose of each ligand relative to a protein. To do so, we first performed the docking simulations using the state-of-the-art software AutoDock Vina, then explored sets of parameters that produced chemically reasonable poses, and finally did the re-scoring using the ComvexPL potential. Later, we critically examined AutoDock Vina sampling method and detected points where it can be improved and also assessed the integration of our inhouse developed ConvexPL scoring algorithm.