APPLICATIONS NEWS

Parallel computing allows scientists to see the biology of cells

2 April 2007



MIT’s Computational and Systems Biology Initiative (CSBi) is using Interactive Supercomputing’s Star-P software to create more complex biological models for drug discovery.

The initiative uses multi-disciplinary teams of biologists, engineers and computer scientists to analyse complex biological phenomena using both computer simulations and lab experiments.

‘Creating the best possible computational models in the shortest time gives us significant predictive power and improves the effectiveness of experimental work,’ said James G. Evans, a CSBi research scientist for imaging and informatics.

In one application, the team will produce predicted microscopic images of cells under different drug treatment conditions. To meet the high demands of such simulations, high performance computers will be used to accelerate the analysis.

Its software previously used a powerful algorithm called K-means to cluster objects in the calculations. Star-P will allow researchers to create parallel K-mean models on their desktop PCs using their existing software, such as Matlab, without needing to be experts in parallel computing.