MIC-Tandem: Parallel X!Tandem Using MIC on Tandem Mass Spectrometry Based Proteomics Data

Abstract

The widespread use of mass spectrometry for protein identification has created an urgent demand for improving computational efficiency of matching mass spectrometry data to protein databases. With the rapid development of chip technology and parallel computing technique, such as multi-core processor, many-core coprocessor and cluster of multi-node, the speed and performance of the major mass spectral search engines are continuously improving. In recent ten years, X!Tandem as a popular and representative open-source program in searching mass spectral has extended several parallel versions and obtains considerable speedups. However, because these parallel strategies are mainly based on cluster of nodes, higher costs (e.g., charge of electricity and maintenance) is needed to get limited speedups. Fortunately, Intel Many Integrated Core (MIC) architecture and Graphics Processing Unit (GPU) are ideal for this problem. In this paper, we present and implement a parallel strategy to X!Tandem using MIC called MIC-Tandem, That shows excellent speedups on commodity hardware and produces the same results as the original program.