Resources
Test to Best: an In Silico Research Center of Excellence
A project to apply computational techniques to cancer research data and generate results, methods and publications that will move cancer treatment forward.
Test to Best, an In Silico Research Center of Excellence, is a joint project between the Translational Genomics Research Institute (TGen) and 5AM Solutions, funded by SAIC-Frederick. The Center's Director is 5AM Chief Science Officer William FitzHugh, its Principal Investigator is Dr. Jeff Kiefer, and its senior advisor is Dr. Michael Berens. The In Silico Research Centers of Excellence have been set up to apply computational techniques to cancer research data and generate results, methods and publications that will move cancer treatment forward.
The Center will collaborate with and analyze data from an already-funded study, the Ivy Genomics-Based Medicine (IGBM) Project. The IGBM Project is using selected human glioblastoma multiforme (GBM) xenografts implanted in mouse brains to test the efficacy of multiple targeted treatment regimens. Genomic profiles from these xenografts will be used to create predictive methods so that genomic data from new patient GBM samples can be used to personalize treatments for those patients.
Primary Goals
This effort will address two primary research goals.
- The first goal is to compare the genomic data from the available xenografts to patient GBM samples from The Cancer Genome Atlas project (TCGA). These comparison results will be used to select a representative subset of xenografts to use for drug treatment studies.
- The second goal is to create predictive models to guide treatment selections for new patient GBM samples. Methods to map new patient samples to the xenograft samples will be developed. These mappings can be used to pick treatments based on the efficacy of the treatment regimens on the closest-matching xenograft.
Computational Methods
Several computational methods will be used to address these goals. Gene Set Enrichment Analysis (GSEA) will be used to analyze GBM genomic data from TCGA project to identify and prioritize high-scoring gene sets. The selected gene sets will be further analyzed using rough sets theory to identify the core signatures from those gene sets. Those core signatures will be used to assess the similarity of the TCGA GBM samples to the xenograft samples. In addition, gene contexts, sets of genes that are highly regulated in a subset of samples but unregulated in others, will be identified in both TCGA and xenograft samples. These contexts can serve as sets for GSEA analysis, as well as used to identify sample subsets that show consistent response or resistance to particular treatment regimens. Lastly, gene sets will be identified that are enriched in responders versus non-responders for each treatment regimen, providing an additional method for predicting efficacy of new patient samples.