Global Associations between Copy Number and Transcript mRNA Microarray Data: An Empirical Study

Abstract

With an increasing number of cancer profiling studies assaying both transcript mRNA and copy number expression levels, a natural question then involves the potential to combine information across the two types of genomic data. In this article, we perform a study to assess the nature of association between the two types of data across several experiments. We report on several interesting findings: 1) global correlation between gene expression and copy number is relatively weak but consistent across studies; 2) there is strong evidence for a cis-dosage effect of copy number on gene expression; 3) segmenting the copy number levels helps to improve correlations.

Keywords

• circular binary segmentation
• high-dimensional data
• machine learning
• two-color microarray platform