Development of a Meta-Analysis Method to Rank Colorectal Cancer Biomarkers

High throughput gene expression profiling involves determining which genes are turned on or off in a particular biological sample.  Since the development of such techniques, many experiments have been performed to determine which genes are expressed in one type of biological sample but not another type.  For example, many high throughput gene expression profiling experiments have been conducted to compare cancerous colon samples versus healthy colon samples.  The genes with variable on or off patterns in one state versus the other may be useful as diagnostic and/or prognostic markers.  However, while many such studies have already been performed, the agreement among studies is often quite low due to the heterogeneity of samples and differences in protocols.  In this current study, we have conducted a meta-analysis of 25 colorectal cancer expression profiling studies.  We demonstrated that the overall level among studies, while low, is highly significant.  The high ranking candidates are currently being validated on independent tissue samples and these results will likely yield useful diagnostic and prognostic biomarkers for colorectal cancer.

Full citation

Chan SK, Griffith OL, Tai IT, Jones SJ.  Meta-analysis of colorectal cancer gene expression profiling studies identifies consistently reported candidate biomarkers.  Cancer Epidemiol Biomarkers Prevention 2008 17(3):543-552.