An Improved Method for Exon Array Data Analysis

Abstract

Alternative splicing of genes is usually induced by some severe diseases. Exon array manufactured by Affymetrix Company has been widely used to detect alternative splicing. Utilizing the mapping between the gray values of splice isoforms and the array probes, we proposed the calculation of gene expression level based on Kseq model. Aiming at the massive amount of array data, an algorithm for parallel computation using multi-core processor was presented. The algorithm was verified through experiments using real datasets and compared with the commonly used algorithm. It was found that parallel computation improved the computation efficiency of the model and the new model enhanced the computation accuracy in subsequent bioanalysis. Keywords: Exon array, gene expression, alternative splicing Parallel Comp