Integrated analysis of gene expression and copy number data using sparse representation based clustering model

Among biological measurements, DNA microarray gene expression and array comparative genomic hybridization (aCGH) have been widely used. Due to the vast information of the biological data, various clustering techniques have been developed to identify subsets of genes with specific gene expression patterns and large variations across samples. Since integrated analysis of genomic data from different sources can further increase the reliability of biological analysis results, methods of integrating and analyzing different types of genomic measurements have emerged. In this work, we jointly examine gene expression and copy number data and iteratively project the data on different clusters through the sparse representation based clustering (SRC) model. Our method has been tested on a breast cancer cell lines data and a breast tumors data set. In addition, simulated data sets were used to test the robustness of the method to noise. Experiments showed that our proposed method can effectively identify genes with large variations in gene expression and copy number, and locate genes that are statistically significant in both measurements. The proposed method can be applicable to a wide variety of biological problems where joint analysis of biological measurements is a common challenge.