Survival analysis for the identified cancer gene subtype from the co-clustering algorithm

Abstract

Cancer gene subtype information is significant for understanding tumour heterogeneity. The early detection of cancer and subsequent treatment can be lifesaving. However, it is hard clinically and computationally to detect cancer and its subtypes in their early stages. Therefore, we extend the analysis and results from Machap et al. (2019), to include the Kaplan-Meier survival analysis with the integration of gene expression and clinical features data. There are two cancer datasets used for the analysis: breast cancer and glioblastoma multiforme. The luminal type was the common subtype of breast cancer, showing a higher survival rate. Whereas the Proneural subtype in glioblastoma multiforme has a little longer survival rate than the other three subtypes. These molecular differences between subtypes have been shown to correlate very well with clinical features and survival parameters to help understand the disease and develop better therapeutic targets.