Trans-study Projection of Genomic Biomarkers in Analysis of Oncogene Deregulation and Breast Cancer

This article focuses on the use of Bayesian concepts and methods in the trans-study projection of genomic biomarkers for the analysis of oncogene deregulation in breast cancer. The objective of the study is to determine the extent to which patterns of gene expression associated with experimentally induced oncogene pathway deregulation can be used to investigate oncogene pathway activity in real human cancers. This is often referred to as the in vitro to in vivo translation problem, which is addressed using Bayesian sparse factor regression analysis for model-based translation and refinement of in vitro generated signatures of oncogene pathway activity into the domain of human breast tumour tissue samples. The article first provides an overview of the role of oncogene pathway deregulation in human cancers before discussing the details of modelling and data analysis. It then considers the findings based on biological evaluation and Bayesian pathway annotation analysis.