Virtual Screening (VS) methods can considerably aid clinical research, predicting how ligands interact with drug targets. Most VS methods suppose a unique binding site for the target, usually derived from the interpretation of the protein crystal structure. However, it has been demonstrated that in many cases, diverse ligands interact with unrelated parts of the target and many VS methods do not take into account this relevant fact.
We present BINDSURF, a novel VS methodology that scans the whole protein surface in order to find new hotspots, where ligands might potentially interact with, and which is implemented in last generation massively parallel GPU hardware, allowing fast processing of large ligand databases.
In this work we have presented the BINDSURF program. We have shown the details of its modular design, so other users can modify it to suit their needs.
In view of the results obtained, we conclude that BINDSURF is an efficient and fast methodology for the unbiased determination on GPUs of protein binding sites depending on the ligand. It can also be used for fast pre-screening of a large ligand database, and its results can guide posterior detailed application of other VS methods. Its utilization can help to improve drug discovery, drug design, repurposing and therefore aid considerably in clinical research.
In the next steps we want to substitute the Monte Carlo minimization algorithm for more efficient optimization alternatives, like the Ant Colony optimization method, which we have already implemented efficiently on GPU and implement also full ligand and receptor flexibility.
Lastly, we are also working on improved scoring functions to include efficiently metals, aromatic interactions, and implicit solvation models.
The program code is available upon authors’ request.
Irene Sánchez-Linares, Horacio Pérez-Sánchez, José M Cecilia and José M García. High-Throughput parallel blind Virtual Screening using BINDSURF. BMC Bioinformatics 2012, 13(Suppl 14):S13 [doi:10.1186/1471-2105-13-S14-S13] [Free PDF]