Team:Groningen/modeling cumulus

Cumulus is the Cloud based modelling and data exchange program that we developed. Its purpose is to show, as a proof of concept, that it is possible to integrate both the exchange of data, the production of comprehensive models, and the sharing of computational resources into one platform.

To this end we developed a new collaborative Biology Framework based on Service Oriented Architectures (SOA) and on a Web Services. This approach simplifies and harmonizes the work carried out on the Cloud and Grid platform by geographically dispersed scientists. The new collaborative SOA Cloud Framework (called Cumulus) provides an efficient and easy way to run distributed software.

This capability is put to use in a parameter optimization program, (currently a genetic algorithm) to find the optimal parameters for submitted models. In other words, Cumulus makes it simple to use cloud computing resources to find parameters for models, and at the same time, leads to the sharing of information.

Below are links to look at the different parts of the Cumulus project.

• Design Cumulus.
• Why not using existing software?.
• Overview of the simulation system.
• Explanation of the parameter optimization algorithm.
• Results with the cumulus system.
• Next version of the cumulus system.