Analysis of large scale gene expression data sets from TCGA identifies potential candidate genes under dosage compensation in breast cancer

Abstract

In a previous effort, we developed a model to study the hypothesis of gene dosage compensation in the limited data set of NCI60 cancer cell lines. Currently, we have developed a pipeline to extract data from TCGA in order to expand our analysis into larger data sets. To validate our approaches we focused on breast cancer as a test case. Using GMM we identified a cluster of candidate genes, which were subjected to a linear fit to determine potential dosage compensation. This adds support to our hypothesis with about dosage compensation in the NCI60 panel. These genes have multiple structural and functional annotations. These include their over-representation in chromosomes 8,11 and 17. Molecular functions such as protein complex interactions were identified. We found that the candidates are interconnected by a large-scale network of miRNA-TF interactions. These results contribute to the understanding of the phenomenon of gene dosage compensation in cancer and its possible application in developing novel therapies against aneuploid cancer.