GPU accelerated molecular docking with parallel genetic algorithm
Kwoh, Chee Keong
Date of Issue2012
IEEE International Conference on Parallel and Distributed Systems (18th : 2012 : Singapore)
School of Computer Engineering
Bioinformatics Research Centre
Molecular docking is a widely used tool in Computer-aided Drug Design and Discovery. Due to the complexity of simulating the chemical events when two molecules interact, highly accelerated molecular docking programs are of great interest and importance for practical use. In this paper, we present a GPU accelerated docking program implemented with CUDA. The hardware-enabled texture interpolation is employed for fast energy evaluation. Two types of parallel genetic algorithms are mapped to the CUDA computing architecture and used for the search of optimal docking result. Comparing to the CPU implementation, the GPU accelerated docking program achieved significant speedup while producing comparable results to the CPU version. The source code is made public at http://code.google.com/p/cudock/.
DRNTU::Engineering::Computer science and engineering
© 2012 IEEE.