Genomic, proteomic and informatics analysis of doping
AIMS OF THE PROJECT. This goal of this project has been to compare global patterns of gene expression as described in disparate WADA-sponsored studies of the effects of doping agents of methods such as erythropoietin or hypoxia, growth factors such as human growth hormone and IGF-1, steroids and others. To achieve that goal, we have continued to refine the informatics infrastructure and have developed protocols for the application of computational methods for large-scale meta analysis of gene expression data sets from three separate and independent doping studies. We approached the directors of a number of WADA-sponsored studies to obtain data bases that could all be subjected to uniform analytical procedures to identify those presumably few common features that might constitute rigorous markers of exposure to doping manipulation. We received extensive data sets from two other WADA-supported investigators – James Rupert of the University of British Columbia and Dr. Tejvir Khurana of the University of Pennsylvania – and have identified preliminary candidate signatures for further validation and comparison with results of additional data sets to be included in future analyses.
We have successfully used the WADA Informatics facility to down-load and analyze several large transcriptomic datasets, including the one generated in our own laboratory for the IGF-1 study (Bhasker and Friedmann, 2008), as well as datasets rom the WADA-supported studies of James Rupert at the University of British Columbia and Dr. Tejvir Khurana of the University of Pennsylvania. The purpose of these preliminary studies has been to identify and solve the up-loading difficulties that outside users might encounter. The results of that exercise are presented in detail in the attached figures. Briefly, we have demonstrated that a comparison of studies using disparate methods of creating hypoxic conditions in mice reveal similar patterns of transcriptional dysregulation, despite many major differences in experimental design. These similarities include established categories of biological processes, molecular function and specific gene aberrations (slides 4-6) of Powerpoint summary. Those similarities may constitute the beginnings of a rudimentary molecular “signature” for metabolic and gene expression responses to hypoxia and/or to possibly related manipulations such as artificially augmented blood production in a sport setting (Slide 7). In contrast, a comparison of hypoxia conditions with the expected “negative control” effects of IGF-1 exposure of muscle stem cells reveals fewer transcriptional changes in common with the hypoxic conditions, as expected. We emphasize that these results require extensive validation and corroboration with other related and unrelated data sets from other WADA investigators. That will be the emphasis for future studies with this system.