Integrating heterogeneous datasets for cancer module identification

The availability of multiple heterogeneous high-throughput datasets provides an enabling resource for cancer systems biology. Types of data include: Gene expression (GE), copy number aberration (CNA), miRNA expression, methylation, and protein-protein Interactions (PPI). One important problem that can potentially be solved using such data is to determine which of the possible pair-wise interactions among genes contributes to a range of cancer-related events, from tumorigenesis to metastasis. It has been shown by various studies that applying integrated knowledge from multi-omics datasets elucidates such complex phenomena with higher statistical significance than using a single type of dataset individually. However, computational methods for processing multiple data types simultaneously are needed. This chapter reviews some of the computational methods that use integrated approaches to find cancer-related modules.