Survival Analysis with the Integration of RNA-Seq and Clinical Data to Identify Breast Cancer Subtype Specific Genes

摘要

Breast cancer is one of the most widespread forms of cancer that affects a significant portion of the female population today. Its early detection and subsequent treatment can be life saving. However, it is difficult clinically and computationally to detect breast cancer and its subtypes in their early stages. On the other hand, Next Generation Sequencing (NGS) techniques have significantly accelerated the process of mapping the human genomes by providing high-throughput expression data of RNA. In this work, we study such NGS based expression data of mRNAs with the clinical data in order to (a) rank the genes based on their importance in survival of breast cancer subtypes and (b) find the relation between the up/down regulation of genes and survival probability of a population. In this regard, first volcano plot is used to find the differentially expressed genes for each subtype, and second, such genes are used to perform the Kaplan-Meier survival analysis with the integration of mRNA expression and clinical data to rank the genes by their importance in survival of breast cancer subtypes. These genes are ranked based on the p-value and significant genes are filtered out by considering the cut-off as p-value < 0.05 for each breast cancer subtype. In our analysis, we have found a relation between gene regulation and survival probability, e.g. up and down regulated genes of a population show low rate of survival of that population. Moreover, for the biological significance, PPI network and KEGG Pathway analysis are conducted on a common set of genes that are present in all subtypes. The datasets, code and supplementary materials of this work are provided online (http://www.nitttrkol.ac. in/indrajit/projects/mrna-survival-breastcancer-subtypes/).