Biostatistics Shared Resource

PROJECT SUMMARY: BIOSTATISTICS SHARED RESOURCE The Biostatistics Shared Resource (BSR) provides collaborative statistical support to Fred Hutch/University of Washington/Seattle Children’s Cancer Consortium members across the Research Programs. Faculty within the BSR are also a part of the Biostatistics and Computational Biology (BCB) Research Program, but activities within the BSR are distinct from those within BCB, as the BSR provides no-cost biostatistical support to Consortium members working on projects that do not have dedicated funding for biostatistical support. In addition, the BSR assists members across all Programs in developing grant proposals. It is expected that such proposals will have biostatistical support built into the budget, in which case – if a grant is awarded – a funded collaboration with a biostatistician would subsequently take place outside the auspices of the BSR. In this sense, the BSR frequently contributes to the spawning of NIH-funded research. In addition, some Consortium members are interested in conducting their own statistical analyses, and BSR members educate these investigators on how to conduct such analyses appropriately; however, we emphasize that a BSR member should vet all analyses to ensure their validity. The BSR comprises seven faculty-level statisticians and three masters-level statisticians, and the CCSG currently funds roughly 1.0 FTE. The level of support for each biostatistician in the BSR ranges from approximately 5-20%. Each biostatistician is primarily funded by research grants and contracts independent of their BSR activities. The BSR also works closely with other CCSG Shared Resources, for example the Genomics & Bioinformatics Shared Resource (G&BSR). Consortium investigators often approach the G&BSR where data from biological samples (such as RNAseq expression data) need to be bioinformatically processed. The G&BSR often generates customized summaries which are shared with the BSR for more specialized statistical analyses. The G&BSR will engage the BSR as needed where BSR statisticians then implement appropriate statistical tests or modeling. The G&BSR also often refers users with specialized needs (e.g., building prediction models or classifiers) directly to the BSR. The G&BSR will also often advise the BSR on interpretation subtleties for assays such as RNA exome profiling. Conversely, the BSR often refers researchers to the G&BSR for help with sequence-based projects such as phasing MHC loci or analyzing long-read sequence data. The CCSG-supported effort for the BSR ensures that a stable staff of highly skilled biostatisticians is available to Consortium investigators as needed.