DSIMBench: A Benchmark for Microarray Data Using R

TL;DR: DSIMBench is proposed, a benchmark containing two classic microarray analysis functions with eight different parallel R workflows, and the benchmark is evaluated in the IC Cloud testbed platform.

Abstract: Parallel computing in R has been widely used to analyse microarray data. We have seen various applications using various data distribution and calculation approaches. Newer data storage systems, such as MySQL Cluster and HBase, have been proposed for R data storage; while the parallel computation frameworks, including MPI and MapReduce, have been applied to R computation. Thus, it is difficult to understand the whole analysis workflows for which the tool kits are suited for a specific environment. In this paper we propose DSIMBench, a benchmark containing two classic microarray analysis functions with eight different parallel R workflows, and evaluate the benchmark in the IC Cloud testbed platform.