Optimization of hadoop cluster for analyzing large-scale sequence data in bioinformatics

Unexpected growth of high-throughput sequencing platforms in recent years impacted virtually all areas of modern biology. However, the ability to produce data continues to outpace the ability to analyze them. Therefore, continuous efforts are also needed to improve bioinformatics applications for a better use of these research opportunities. Due to the complexity and diversity of metagenomics data, it has been a major challenging field of bioinformatics. Sequence-based identification methods such as using DNA signature (unique k-mer) are the most recent popular methods of real-time analysis of raw sequencing data. DNA signature discovery is compute-intensive and time-consuming. Hadoop, the application of parallel and distributed computing is one of the popular applications for the analysis of large scale data in bioinformatics. Optimization of the time-consumption and computational resource usages such as CPU consumption and memory usage are the main goals of this paper, along with the management of the Hadoop cluster nodes.