A new supercomputing workload management system is aiding scientific work by Cancer Research and the Cambridge Research Institute's human genome project.
Cancer Research UK is using Platform Computing's LSF (Load Sharing Facility) software to improve cluster efficiency and reduce IT costs on the CRI genome research.
By integrating Platform LSF with a new advanced genetic sequencing platform, the institute has already gained greater insight into genetic cancer mutations that will lead to scientific breakthroughs in the areas of cancer diagnosis, treatment and prevention, said Cancer Research.
“Platform LSF gives us the means to produce and manage a wealth of gene sequencing data that we could only have dreamed about previously,” said Peter Maccallum, head of IT and scientific computing at Cancer Research UK in Cambridge. “This has already lead to tangible published work looking into breast cancer, and is proving its worth in helping our researchers further the understanding of how cancers progress.”
Cancer Research has reportedly doubled the processing power of its centralised computer cluster over the last three months, due to the increasingly data-driven nature of its work. Prior to implementing Platform LSF, CRI’s 21 research groups employed separate computing resources in separate locations, which drove up server costs, reduced utilisation rates and increased server maintenance.
By orchestrating workloads and managing CRI’s research applications in a single data centre, Platform LSF has enabled CRI to save approximately £50,000 by removing hardware and maintenance duplication across each location, while increasing the amount of data processed. Cancer Research says the institute can now direct more computing resources directly to its research teams "to use in a more timely and cost efficient manner".
CRI has already saved the equivalent in man hours of one full-time employee by integrating Platform LSF, says Cancer Research. As a result, the institute plans to scale Platform LSF internally by adding more servers as compute requirements increase.
CRI is also collaborating with Platform Computing to architecturally support cross-organisation systems for HPC (high performance computing) clusters, that will enable CRI to collaborate with other research organisations in order to meet the growing demand for genomics research.
In other recent medical technology news, scientists at Cambridge University are developing a computer system that can read vast amounts of scientific literature, make rapid connections between facts and develop hypotheses. Cambridge University said most biomedical scientists cannot keep on top of reading all of the publications in their field, let alone an adjacent field. As a first step to solving the problem, Cambridge has developed its CRAB text-mining tool.