PBS Professional at Toulouse Genopole: Accelerating Genomics Research

• Infrastructure as a Service (IaaS) - Cloud Computing
• Platform as a Service (PaaS) - Application Development Platforms

• Electrical Grids
• Life Sciences

• Product Research & Development

• Smart City Operations

• System Integration

The Toulouse Midi-Pyrénées Genopole, a research program established in 1999 in southern France, was facing a significant challenge. The program, which is part of the National Genopole Network, was dealing with an increasing number of complete genome sequences that required more processors and intensive utilization. The initial compute platform for bioinformatics at the Toulouse Genopole was a four-processor Dell server, later supplemented by a four-CPU IBM system. However, these resources were insufficient for the ambitious bioinformatics program and the researchers developing bioinformatics tools in three Genopole-connected laboratories. The demand for computational resources was steadily increasing, and the existing infrastructure was unable to keep up.

The Toulouse Midi-Pyrénées Genopole is a research program set up in 1999 in southern France as part of the National Genopole Network. The initiative has drawn together a mix of public laboratories, biotech companies, and educational institutions at seven locations across France. The Toulouse Genopole comprises thirty academic laboratories, with some 3,000 researchers working in five basic fields: Microbiology, plants and animals, health, ethics, and bioinformatics. The Genopole’s bioinformatics resources are hosted by the Toulouse INRA (National Agronomic Research Institute) computing center. The Genopole serves four kinds of users: Biologists who access bioinformatics tools through the platform’s open website, UNIX-literate biologists who run bioinformatics tools in console mode, software engineers developing new interfaces and workflow systems for large-scale genomics or bioinformatics programs, and scientists developing new algorithms and methodologies for gene detection, comparative genomics, microarray analysis, sequence searches and other fields.

To address this challenge, Altair provided a powerful workload manager with Job Arrays to increase the power and user-friendliness of Toulouse Genopole's systems. Job Arrays, a feature of Altair's PBS Professional workload management software, allows users to submit and manage a group of jobs as a single job. This feature is particularly useful in large-scale bioinformatics studies. The French Altair team worked with the Genopole to integrate their bioinformatics applications with the Job Arrays function. The software was installed on INRA’s LNXI Opteron cluster when it was upgraded to 96 processors and came online in January 2006. Looking ahead, INRA has implemented web-based services into its HPC platform at Toulouse Genopole and is attracting new visitors. With help from Altair, the web interface will soon be integrated with PBS Professional to manage job submission from remote users.

• The implementation of Altair's PBS Professional workload management software with Job Arrays has significantly increased the power and user-friendliness of Toulouse Genopole's systems. This has enabled scientists to write genomic sequence analysis workflows much more easily. The upgraded cluster went immediately into beta operation with three users, demonstrating the efficiency and effectiveness of the solution. Furthermore, the integration of web-based services into the HPC platform at Toulouse Genopole is attracting new visitors, extending the reach and impact of the Genopole's research. With the upcoming integration of the web interface with PBS Professional, the Genopole will be able to manage job submission from remote users, further enhancing its operational capabilities.

• Increased the number of processors from 4 to 96
• The platform’s open website attracts 15,000 visits in a typical month
• Around 100 UNIX-literate biologists run bioinformatics tools in console mode