Integrated analysis of DNA copy number and gene expression microarray data using gene sets

Renée X Menezes,Judith M Boer,Marten Boetzer,Melle Sieswerda,Gert-Jan B Van Ommen

doi:10.1186/1471-2105-10-203

Abstract

BackgroundGenes that play an important role in tumorigenesis are expected to show association between DNA copy number and RNA expression. Optimal power to find such associations can only be achieved if analysing copy number and gene expression jointly. Furthermore, some copy number changes extend over larger chromosomal regions affecting the expression levels of multiple resident genes.ResultsWe propose to analyse copy number and expression array data using gene sets, rather than individual genes. The proposed model is robust and sensitive. We re-analysed two publicly available datasets as illustration. These two independent breast cancer datasets yielded similar patterns of association between gene dosage and gene expression levels, in spite of different platforms having been used. Our comparisons show a clear advantage to using sets of genes' expressions to detect associations with long-spanning, low-amplitude copy number aberrations. In addition, our model allows for using additional explanatory variables and does not require mapping between copy number and expression probes.ConclusionWe developed a general and flexible tool for integration of multiple microarray data sets, and showed how the identification of genes whose expression is affected by copy number aberrations provides a powerful approach to prioritize putative targets for functional validation.

Highlights

Genes that play an important role in tumorigenesis are expected to show association between DNA copy number and RNA expression
Gene dosage changes play an important role in tumor development; oncogenes may be enhanced by DNA amplification and tumor suppressor genes may be inactivated by a physical deletion
We evaluate the results by producing receiver-operating characteristic (ROC) curves of the regional model for each case, and consider that an effect is detectable if there is power of at least 60% to detect it using an FDR of 10%

Summary

Introduction

Genes that play an important role in tumorigenesis are expected to show association between DNA copy number and RNA expression. Optimal power to find such associations can only be achieved if analysing copy number and gene expression jointly. Some copy number changes extend over larger chromosomal regions affecting the expression levels of multiple resident genes. Tumor cells accumulate genetic damage, including changes in DNA copy number, sequence and methylation, resulting in the dysfunctioning of key regulators [1]. The advent of microarray technology has allowed genome-wide monitoring of these molecular changes at the DNA and RNA level. High-resolution arraybased comparative genomic hybridization (array-CGH) has allowed the delineation of recurrent DNA copy number alterations in tumors [8,9,10]. Integrated analysis of both copy (page number not for citation purposes)

Methods

Results

Discussion

Conclusion