Poster Sessions

BIOINFO-16

Calvin Johnson

C. Johnson, T. Caruso, R. Martino, S. Bailey, D. Russ, B. Trus, H. Cheung

IRP Grid: A Feasibility Study of Grid Technologies for the NIH Intramural Research Program

Basic and translational research efforts often involve the cooperation of many centers from multiple locations around the world that need to exchange terabytes or more of data generated by advanced technologies. These huge datasets exist in various forms, such as images and high-throughput genomic data (including numerous whole-genome and population-based studies), for which computer-intensive analysis is required. Currently, although the NIH intramural research program includes many high-end computing resources, most investigators utilize only a small subset of these facilities. Even when resources are available, investigators face steep learning curves or must acquire (or pay for) niche expertise in order to take advantage of these resources. Furthermore, data sharing-facilities are limited to small communities at NIH. We propose developing a grid environment for the NIH Intramural Research Program, called the IRPgrid to support future biomedical research endeavors and to promote resource sharing. By maintaining compliance with caBIG and TeraGrid, the IRPgrid will be able to take advantage of previous investments in grid computing. The IRPgrid is envisioned to provide the entire NIH intramural program with grid-enabled high-performance computing, massive storage capability, and semantically interoperable software applications.