Identification genes of colorectal cancer with integrated data via Block-sparse NMFL21

摘要

Colorectal cancer is the second leading cause of cancer deaths in the word. The assessment of the colorectal cancer is determined by the clinical stage which denotes the number of lymph nodes with metastatic cancer and the identification of metastatic sites in other organs. While the cancer genes associated with colorectal cancer have been characterized, less is known about the identity of special cancer genes that are responsible for advanced clinical stage and metastasis. To characterize genetic aberration in colorectal cancer, we conducted a Block-sparse nonnegative matrix factorization with L_(2l) -norm (Block-sparse NMF L_(21)) analysis of multiple genomic data which include mutation, DNA copy number, promoter methylation, messenger RNA and microRNA expression from The Cancer Genome Atlas (TCGA). With the analysis of the method, we ranked the genes by the scores and identified a set of top gene predictors of advanced clinical stage. The results demonstrate that the analysis of integrated data through block-sparse NMF L_(21) is more comprehensive.