|
Abstract:
|
High-dimensional omic data derived from different technological platforms have been extensively used to facilitate comprehensive understanding of disease mechanisms. It is also recognized that incorporation of some biological information (e.g. pathway) in the analysis of omic data can lead to more accurate and interpretable results. We propose a statistical framework of shared informative factor models that can jointly analyze multi-platform omic data, explore their associations with a disease phenotype, and incorporate pathway information while integration. Extensive simulation studies demonstrate the performance of the proposed method in terms of biomarker detection and prediction accuracy. We also illustrate the applicability of the proposed method using a TCGA kidney cancer data set. The association of mRNA expression and protein expression with survival of kidney cancer patients is explored.
|
