CAMBRIDGE-CRANFIELD HIGH PERFORMANCE COMPUTING FACILITY

Abstracts for Talks at CCHPCF Launch Event
Thursday 30th October 2-6pm, CMS

Systematic Identification of Large Numbers Frog Genes by Comparison with other Organisms
Dr Mike Gilchrist Head of Bioinformatics Group, the Wellcome Institute

Access to the SunFire Galaxy class supercomputer at the HPCF enables individual researchers to do search and comparison tasks for proteins at a scale and rate formally only feasible at sequencing resource centres such as the Sanger Institute. We have recently elucidated a large part of the frog (Xenopus tropicalis) proteome (genes expressed as proteins) in key early embryonic development stages. Part of this requires that we compare the resultant 35,000 or so sequences against all the known proteins in the public databases (approx. 1.5 million). This enables as both to make tentative identification of genes based on degrees of similarity, and to ascertain whether our physical reagents (clones) contain the whole coding sequence for some or all of the identified genes. Whilst this task is possible on (say) a single 2.0 GHz processor CPU it would take weeks to finish the required search, on the SunFire machine running on a single 96 CPU node it takes approximately 3 hours. This is of enormous benefit to us.