Predicting chromosome instability in cancer using gene expression

Predicting chromosome instability in cancer using gene expression data and location information Bioinformatics for Genomic Medicine 2006 04 -24 Do Kyoon Kim

Introduction • Microarray data – High throughput data contains many information • Integration of data & knowledge – Better outcome?

Problem definition • Detect gene expression pattern by chromosome instability in cancer using microarray gene expression data and location information (genes within the chromosomal region) • Add additional function to the Chromo. Viz for visualization if possible

Method 1. Clustering each samples by number of chromosome unit - Chromosome unit: genes in user defined distance on chromosome Normal patient A cancer B cancer C cancer

Method 2. Test membership similarity between chromosome unit and cluster using hypergeometric distribution Normal patient 1 3 A cancer 2 4 B cancer 1 2 3 C cancer 1 2 3 4 1 2 4 4 3

Method 3. Compare to each graphs or visualize expression data onto chromosome with associated annotation of location information