|
Abstract:
|
In recent years, cancer molecular classification based on microarray gene expression profiling has led to discovery of novel tumor subtypes with prognostic implications. However, the genetic complexity of a tumor manifests at multiple levels including genomic, epigenomic and transcriptomic alterations. Using glioblastoma, ovarian, and lung cancer as pilot study models, the Cancer Genome Atlas project is generating multiple types of high-dimensional "omic" data to catalog somatic mutations, DNA copy number alterations, DNA methylation patterns, messenger RNA gene expression and microRNA regulation in hundreds of tumors. In contrast to the accumulating wealth of multidimensional data sources, there is a lack of effective statistical and bioinformatic methods for integrative data analysis. In this study, we aim to develop a new integrative paradigm for tumor subtype discovery.
|
