Getting poster data...
Sarah-Jane Schramm1, Anna E Campain1, Simone S Li2, David C Y Fung2, Vivek Jayaswal1, Chi Nam Ignatius Pang2, Richard A Scolyer1&3&4, Yee Hwa Yang1, Marc R Wilkins2 and Graham J Mann1&3 (1The University of Sydney, Sydney, Australia; 2University of New South Wales, Sydney, Australia; 3Melanoma Institute Australia, Sydney Australia; 4Royal Prince Alfred Hospital, Sydney, Australia )In melanoma there is an urgent need to identify novel biomarkers that can aid in predicting patient outcome. Cancer systems biology has ushered in new methods that may be used for this purpose. Here we used one such approach to generate networks of molecular interactions from publicly available gene expression microarray data of human melanoma tissue. Specifically, we tested for significant differences in the average correlation of gene expression between highly connected genes and their direct, physical interaction partners (at the protein level) among patients with either good or poor outcome. For all datasets tested, we observed significantly altered dynamics in molecular networks between patient groups. We are using network visualisation to further characterise these networks e.g. to identify potential candidate mutations that drive the prognostic footprint. This includes, for the first time, overlaying the results of similarly evaluated miRNA expression data from matched tumor samples. We are currently evaluating the networks as prognostic tests and are encouraged by preliminary results indicating that patient outcome may be predicted with error rates of approximately 0.25.