Identification of related genes with survival in renal carcinoma by using lassoed principal components method

Identification of correlated genes with survival by gene expression data is an important application of microarray data. The purpose of this study is to identify correlated genes with survival of conventional renal cell carcinoma (cRCC) patients based on gene expression profiles. This study is a survival analysis with high dimensional covariates and containing 14814 gene expression measurements from 177 patients with cRCC. Lassoed principal components (LPC) method is used for identification associated genes with survival. LPC score uses information of all of gene expressions for computation a gene score. Finally False Discovery Rate (FDR) method is used to identify significant genes. Statistical analysis is done with using the R software. The lowest error is satisfied with using the cutoff 0.001 for FDR criteria and with studying 1041 related genes with survival of cRCC patients. 11 genes are identified as most significant genes with survival of cRCC patients, after ranking the genes with their LPC scores with regard to their differentially expressions. The LPC scores of these 11 genes are negative, so increase of these gene expressions are related to increase of the survival of cRCC patients and in the other words the increase of these gene expressions are protective factors in cRCC patients.