Large-scale Computer Simulations Reveal New Insights Into Antibiotic Resistance

Experiments to find out how changes at the molecular level arecausing this resistance are difficult and, so far, have not been done.Now, however, Peter Coveney and co-workers from UCL and Queen Mary,University of London have investigated the problem using computermodelling techniques. Their findings are amongst several outputs of theUK e-Science programme that are discussed in a special Theme Issue ofPhilosophical Transactions of the Royal Society A* which is publishedon 15 August.

They took experimental data gathered from other organisms tobuild computer models of the sites where drug molecules interact withan organism's protein molecules. They then ran simulations andvisualised what happens when a drug molecule approaches each site forboth normal and drug-resistant strains of S. pneumoniae.

The simulations and visualisations exploited highly scalableparallel code running on the UK's national supercomputing facilities."Without the use of e-Science methods, they would have taken months toperform and quite probably would never have been done. With these newmethods, each simulation took just 12 hours," says Professor Coveney.So far, life scientists have had limited access to and interest in suchhigh performance computing resources; with Grid computing, theseresources are becoming more readily accessible.

Professor Coveney and colleagues could see that a very small,but subtle, difference in structure between the normal and drugresistant strains was to blame for the drug resistance. In the normalstrain, a drug molecule binds tightly to the site, but in the drugresistant strain it approaches and then drifts slowly away. If theresults of the simulation are borne out by experiment, they could pointthe way to new drugs to combat disease.

*Large-scale moleculardynamics simulation of native and mutant dihydropteroatesynthase-sulfanimide complexes suggests the basis of dihydropteroatesynthase drug resistance by F.Giordanetto, P. W. Fowler, M Saqi and P.V. Coveney Philosophical Transactions of the Royal Society A 363 1833(15 August 2005)http://www.pubs.royalsoc.ac.uk/phil_trans_phys_scientificgrid.shtml

Notes:

1. e-Science is the very large scale science that can be carriedout by pooling access to very large digital data collections, verylarge scale computing resources and high performance visualisation heldat different sites.

2. A computing grid refers to geographically dispersed computingresources that are linked together by software known as middleware sothat the resources can be shared. The vision is to provide computingresources to the consumer in a similar way to the electric power grid.The consumer can access electric or computing power without knowingwhich power station or computer it is coming from.

3. The UK e-Science Programme is a coordinated £230M initiative involvingall the Research Councils and the Department of Trade and Industry. Ithas also leveraged industrial investment of £30M. The Engineering andPhysical Sciences Research Council manages the Core e-ScienceProgramme, which is developing generic technologies, on behalf of allthe Research Councils.

4. The UK e-Science Programme as a whole is fostering the developmentof IT and grid technologies to enable new ways of doing faster, betteror different research, with the aim of establishing a sustainable,national e-infrastructure for research and innovation. Furtherinformation at http://www.rcuk.ac.uk/escience/