Core 3: Biostatistics and Bioinformatics Core

PROJECT SUMMARY – CORE 3 The principal objective of the Biostatistics and Bioinformatics Core (Core 3) is to provide project investigators a centralized resource for biostatistics and bioinformatics expertise. Statistical issues will be addressed at all levels of investigation: from the design of experiments to the maintenance of data quality; and from conclusions based on formal hypothesis testing, to important leads discovered by thorough data exploration. The specific aims of the Core are: 1.1. Collaborate with project investigators in the formulation of unambiguous hypotheses and hypothesis testing strategies, and in the design of experiments. Biostatistics collaboration will assist researchers at the design stage by: formulating biological and clinical questions as testable statistical hypotheses; devising efficient experimental designs and matching these with appropriate statistical models for both basic and high dimensional informatics data; determining sample sizes necessary to ensure high power; and strongly controlling type I error. 1.2. Provide support for all projects with: formal hypothesis tests in experimental and clinical data that ensure strong conclusions; exploratory bioinformatics analyses that lead to further experiments, refined hypotheses, or discoveries; statistical modeling and sensitivity analyses of complex data; and visual displays of data that clarify conclusions and uncover leads. Core 3 statisticians/informaticians have worked extensively with investigators to develop comprehensive analysis plans that distinguish between formal testing (which provides evidence for strong conclusions) and exploratory bioinformatics analyses (which do not allow for strong conclusions but lay the groundwork for important discoveries). Core statisticians/informaticians will employ the most robust methods for modeling and sensitivity analyses to ensure reliability of results. For all data, visual displays will be emphasized to reveal the reliability of conclusions and aid in the discovery of new leads. 1.3. Provide data management and integration services that ensure high integrity, security and investigator accessibility. Ensure statistical and informatics analysis consistency and transparency across projects, and have a centralized repository of all analysis procedures and results. To promote data integration across projects, Core 3 will have a centralized repository of all data that is secure and accessible to the investigators. Data will be cleaned, merged and de-identified. All missing and outlying data will be reported to investigators. Core 3 will ensure consistency of analysis plans and reports for the same data types across projects. 1.4. Conduct innovative statistical and bioinformatics modeling to directly combat complex data or design problems that arise in projects and where current methods do not apply. When faced with specific issues that arise in these projects where current analysis methods do not apply, Core 3 personnel will either adjust available methods or develop novel procedures to address the problems.