Speakers
Description
Impact
Based on the experience gained with the WISDOM environment, the platform makes an efficient use of the grid services. Thanks to the AMGA metadata catalogue it uses a global information system to share all the biological data used by its services for better interoperability. The system is based on the use of agents, or software wrappers; the agents are designed to run any software integrated in the platform and pick up tasks on the fly. All the information concerning the tasks is stored in the platform information system so input data are simply collected on the WN and results are saved back on the storage elements. The jobs also store some results and statistics dynamically in the platform information system in AMGA. The automatic replication and update service is based on the grid data management system, using LFC and storage elements to manage the physical files and manage the metadata in AMGA.
Justification for delivering demo and technical requirements (ONLY for demonstrations)
In this demonstration, we will show how the different tools can be simply managed through external web pages for simple invocation, or in a more complex way through workflow engines (MOTEUR, Taverna) while monitoring in real time what is actually happening on the grid. The demonstration will focus on user-friendliness and versatility of the platform as its tools can be used transparently and efficiently. It will show on a single screen some use cases, and show how the platform can be used.
Detailed analysis
The environment has been developed with one idea in mind: addressing the needs of the bioinformatics community that works with workflows and that may need the computing power of the grid or just simple computers to run their tasks. By the use of agents which preallocate the resources it allows the quick execution of simulations and makes the grid useful and efficient for short jobs and large-scale deployments alike. The use of an interoperable technology also enhances the integration capabilities of new software and databases, that can easily be used altogether through workflow engines. This also allows to wrap middleware specific aspects to run jobs on multiple grid infrastructures. Advanced tools used in the platform manage automatically the job submissions, replication and update of the biological databases so end-users just have to focus on their workflows and not on the technology.
URL for further information
http://amga02.lpc-rd.fr/plate-forme-bioInfo
Keywords
WISDOM, Production environment, Interoperability, Workload Management, Data Management
Conclusions and Future Work
The WISDOM production environment has evolved to allow not only 'in silico' docking but also other types of application, for instance a data challenge on corn genome sequencing has been done also using the same environment. The environment will also be used, in the near future, to manage the jobs of the HOPE platform. We also plan some specific work to reinforce security and a more transparent integration with future versions of the MOTEUR workflow engine.
