|
Abstract:
|
As the amount of genetic, genomic, proteomic, and phenomic data is increasing, complexity of data management and analysis are also increasing. At the CCTS, we are approaching the new challenges by (1) extending u-scores to multivariate data, including specific partial orderings (e.g., for genomic data), (2) building a grid of PCs to handle whole genome scans, and (3) generating centralized, secure, and Web accessible data bases from a interactive graphical description of the research design. We are now extending this Web-based system (WISDOM) to supporting our clinical investigators with statistical analyses. In assisted mode (confirmatory analyses), statistical analyses are essentially automated, based upon the knowledge about the primary aims. In expert mode (exploratory analyses), the same knowledge base is used to provide study-specific defaults.
|
