GPU-accelerated protein sequence alignment for Jamu prediction

摘要

In-silico analysis is an important step for discovering new drugs. In this work, we propose an implementation of GPU-accelerated Smith-Waterman local alignment algorithm to compute the similarity score of protein that used for drug- target analysis on Ijah Webserver, a web-based tool that provides searching and predicting services for the interaction between plants and diseases. We implement the proposed method in NVIDIA CUDA based GPU's. Performance evaluation of parallel proposed algorithm is measured by calculating the speedup. Based on the performance evaluation, our proposed method is faster than CPU codes with 3.42 times speedup.