In bioinformatics, one of the gold-standard algorithms to compute the optimal similarity score between sequences in a sequence database searches is Smith-Waterman algorithm that uses dynamic programming. This algorithm has a quadratic time complexity which requires a long computation time for large-sized data. In this issue, parallel computing is essential for sequence database searches in order to reduce the running time and to increase the performance. In this paper, we discuss the parallel implementation performance of Smith-Waterman algorithm in GPU using CUDA C programming language with NVCC compiler on Linux environment. Furthermore, we assess the performance analysis using three parallelization models, including Inter-task Parallelization, Intratask Parallelization, and a combination of both models. Based on the simulation results, a combination of both models has better performance than the others. In addition the parallelization using combination of both models achieves an average speed-up of 313× and an average efficiency with a factor of 0.93.
|Number of pages||9|
|Journal||Journal of Next Generation Information Technology|
|Publication status||Published - 1 Jan 2014|
- CUDA GPU computing
- Protein database searches
- Smith-waterman sequence alignment