Efficient Bioinformatics Computations through GPU Accelerated Web Services
share
Web services have become a common means of addressing distributed task execution. There are several such services available at present that cater to the requirements in the domain of bioinformatics. However, it is challenging to find efforts made to utilize Graphics Processing Unit (GPU) accelerated binaries to optimize web services for bioinformatics analyses. Most of the bioinformatics related algorithms are compute-intensive due to their exponential time complexities. Corresponding execution times are further increased as the sequence databases grow rapidly. However, harnessing parallel computational power gives us a pathway to execute these algorithms and allows us to obtain results from web services more efficiently. The work presented in this paper demonstrates the utilization of web services to provide GPU accelerated bioinformatics computations for analytical purposes. A unified service platform is developed in the form of a collection of web services. It utilizes both GPU and CPU processing in order to perform compute-intensive tasks related to biological data analysis. Furthermore, the evaluation results show gains in speed over three folds with improving performance as the length of queries increases.
READ FULL TEXT