Bioinformatic analysis of merged gene expression microarray datasets in rat models of colorectal cancer

RIS ID

73232

Publication Details

Shaw, J., Kerr, C. A., Clarke, J., Lockett, T. and Head, R. 2009, 'Bioinformatic analysis of merged gene expression microarray datasets in rat models of colorectal cancer', The 9th Annual Australasian Microarray and Associated Technologies (AMATA) Conference: "High Throughput Genomic Technologies", AMATA, Australia,

Additional Publication Information

Handbook freely available via website

Abstract

Azoxymethane (AOM) induced carcinogenesis in rats is a central model used in the study of colorectal cancer (CRC). This analysis merged in silico gene expression data from two AOM studies that used different generations of Affymetrix GeneChip® Rat arrays. The first sampled rat colonic epithelium immediately after AOM injection (Acute model, Genome 230 2.0 Arrays), the second sampled colonic epithelium and AOM-induced tumours (Cancer model, 27 weeks after AOM exposure, Gene 1.0 ST arrays). Saline injected controls were included in both studies. The aim of this analysis was to compare the acute and long-term effect of AOM in the carcinogenesis process by merging these two datasets from different generation arrays.

The merger and gene expression analysis was performed using Partek® and pathway and network analysis using Ingenuity® Systems. Genes unique to either array were first filtered by Gene Symbol annotated probe-sets and then by CEL file. The resulting data was then normalised with RMA using the quantile distribution of the Acute arrays. As the Gene 1.0 ST array covers only well-annotated content, the resulting merged list consisted of 12,435 genes. Of this list, 11,437 (92%) were differentially expressed between the Acute and Cancer studies regardless of treatment. This large ‘experimental effect’ due to chip and/or ‘rat’ (e.g. age) differences was then modelled and the residuals analysed for differential expression between treatment groups.

Lists of differentially expressed genes comparing AOM treated rats to normal in proximal and distal colon epithelium were generated using an FDR of 0.05. These were then analysed for affected Functions and Pathways. Comparison analyses between the acute and cancer models found quite different Functions and Pathways involved in the acute compared to the long-term response of the colon epithelium to the AOM insult.

Link to publisher version (URL)

AMATA Conference

Please refer to publisher version or contact your library.

Share

COinS