Bioinformatics

Duke University, Durham NC

Investigators

Linked publications, trials & patents

Abstract

The Bioinformatics shared resource (BiSR), serves as a centralized resource for expertise in applied and theoretical cancer bioinformatics, genomics, computational biology, mathematical modeling, machine learning, single-cell and spatial transcriptomics, and statistical genetics. The BiSR supports DCI members across the continuum of research, including experimental and statistical design for genomic studies, complex genomic data management, integration of broad data sets, computing and statistical analysis, and machine learning. The BiSR provides support for investigator-generated data as well as retrospective data from research databases (e.g., GDC which includes large datasets like TCGA, or dbGaP). The resource’s mission is to provide high-quality, service-oriented, coordinated, and cost-efficient bioinformatics support and infrastructure for DCI members. Emphasis is placed on facilitating increased collaborations across all seven DCI research programs. The mission of the BiSR is accomplished within the framework of adherence to three principles: 1. sound data provenance and statistical principles, 2. literate programming, and 3. reproducible analysis. The BiSR develops and utilizes standardized genomic analysis pipelines (e.g., germline, tumor, and cell-free DNA-Seq, bulk and single-cell RNA-Seq, ATAC- and ChIP-Seq, sequencing of T and B cell receptors (TCR/BCR), and shotgun and 16S bacterial sequencing). These pipelines are constructed on the basis of state-of-the-art published tools, maintained under strict source code version control, and executed within containerized software environments so as to extensible and scalable on local servers, university clusters, and cloud computing environments such as Amazon Web Services (AWS) and Microsoft Azure. It is the philosophy of the BiSR that the scope of scientific discovery and rigor should neither be limited nor compromised due to lack of appropriate and adequate statistical methodology or computational tools. When needed and appropriate, the faculty and staff of the shared resource are able to extend or revise existing methods or develop de novo computational tools to enable DCI members to address scientific questions with requisite rigor and efficiency. The BiSR provides extensive support for writing of scientific abstracts, manuscripts, and grant proposals. The BiSR also serves as a liaison and facilitator between DCI members and other DCI shared resources including BSR, SGT, MCF, DIPC, FG and BRPC. The research support, services, and resources of the BiSR are provided exclusively to DCI members. In the last reporting period, the BiSR provided services to 90 DCI members across all seven DCI Research Programs, accounting for 100% of total usage. Use of this shared resource by DCI members contributed to 166 publications over the project period, 34 of which were in high impact journals, demonstrating the value of services offered by the resource.