The Application of Distributed Computing to the Investigation of Protein Conformational Change
Christopher J. Woods, Jeremy Graham Frey, Jonathan W. Essex · ePrints Soton (University of Southampton) · 2004
Distributed computing is a potentially very powerful approach for accessing large amounts of computational power. Under the umbrella of the comb-e-chem project we have examined distributed computing software and applied it to the problem of investigating protein conformational change. These investigations required the development of protein simulations that were suited to distributed computing. Each simulation was split into many coupled, parallel parts. These proved challenging to schedule on the flexible and unreliable distributed computing resource. Scheduling algorithms were thus written that identified which parts of the simulation were likely to impact the overall efficiency. These parts were then rescheduled to be ‘caught-up ’ via a fast and dedicated cluster. 1. Background Distributed computing is a potentially powerful approach for accessing large amounts of computational power. Cycle stealers, which allow a PC user to donate the spare power of their computer, are now used in a wide range of scientific projects, e.g. the SETI@home study, 1 the CAN-DDO cancer screening project 2 and