Microarray profiling of inflammatory bowel disease
In this study of inflammatory bowel diseases (IBD, i.e. Crohn's disease and ulcerative colitis), the gene transcription profile of colonic IBD resection specimens were analysed by oligonucleotide microarray analysis. A total of 33,625 genes were profiled across 23 colonic mucosa samples; 5 involved Crohn's disease, 4 uninvolved Crohn's disease, 5 involved ulcerative colitis, 3 uninvolved ulcerative colitis and 6 samples from macroscopically normal areas of colorectal cancer resections (controls). A number of data-mining tools, encompassing clustering (e.g. hierarchical & K-means) and matrix-based methods were evaluated for the analysis of this microarray data. Mining strategies were formulated, tested and then applied to the data set to identify genes showing interesting and novel expression patterns across the samples. The application of these tools to the data set resulted in the generation of gene expression profiles for Crohn's disease and ulcerative colitis. Genes of interest were annotated using publicly accessible sequence and literature databases. Potential links by previous research in the inflammatory bowel disease field were analysed for selected genes. Common pathways emerging from the annotation effort and potentially linking together several of the genes of interest were investigated. Specifically, the energy deficiency hypothesis proposed by Roediger in 1980 and the relevance of potential cancer and apoptosis related genes were reviewed with regard to the findings.