caGrid in Action

Featured Project: TROPIX

Biomedical researchers at the University of Minnesota and at the Mayo Clinic are seeking ways to allow collaborating research groups to efficiently share datasets and data analysis tools. Such capabilities would enable researchers to improve the quality of their work and foster new collaborative research and clinical projects. In 2007, the Minnesota Supercomputing Institute (MSI) and the Bioinformatics group at the Mayo Clinic received a 2-year grant from the Minnesota Partnership for Biotechnology and Medical Genomics to develop a system to provide these capabilities. This project is called TROPIX.

TROPIX is MSI's first open source project based on the caGrid software infrastructure. MSI is also developing other projects (ProTIP and the TINT Portal) that heavily utilize the software and infrastructure that was developed during TROPIX. Each project provides unique and complementary features to enable research:

By using the caGrid infrastructure as a base, MSI and the Mayo Clinic were able to quickly provide useful tools for the research community at the University of Minnesota. Over the last 9 months staff at the Center for Mass Spectrometry and Proteomics have already used the TROPIX, ProTIP and TINT Portal software to analyze data for 17 PIs in 11 departments at the University of Minnesota.

"I operate a mass spectrometer for the core facility. TROPIX is THE way I analyze post-run data."

• Dr. Matt Stone, Postdoctoral Associate at the Center for Mass Spectrometry and Proteomics

"TROPIX has provided essential and enabling infrastructure for numerous projects using mass spectrometry-based proteomics in the Griffin lab."

• Dr. Tim Griffin, Department of Biochemistry, Molecular Biology, and Biophysics

Figure 1. MSI's four caGrid based projects and how they interact with the MSI-Mayo private grid.

In order to ensure secure and controlled access to resources, MSI and the Mayo Clinic have utilized the caGrid GAARDS infrastructure along with traditional firewall techniques to setup a private federated Grid between the two institutions (see Figure 1). Both institutions have deployed the caGrid Authentication and Dorian services to provide grid credentials for local LDAP users. This allows researchers at University of Minnesota to use their institutional accounts and credentials to access the Minnesota-Mayo private Grid and remotely access caGrid-enabled resources at each institution. Because caGrid uses the same X509 credentials Globus uses, MSI has developed analytical services that can use users' grid credentials to submit jobs to its clusters and supercomputers.

So far, MSI has created and deployed over a dozen services onto the production UMN/Mayo grid. As the number of services has grown MSI has utilized caGrid tools to improve the process of maintaining their services.

"The caGrid Introduce toolkit provides a feature-rich graphical interface for developing and deploying caGrid services. Introduce's modular design has enabled MSI to create extensions to develop and deploy services directly from the command line. This drastically reduces the overhead involved when trying to maintain a large number of services."

• Mark Nelson, Lead Software Developer at the Minnesota Supercomputing Institute

The caGrid infrastructure has provided an essential base for the development of software and tools at MSI. Researchers are already using MSI's caGrid based services to accomplish cutting edge research. By continuing to grow and improve the number of services it offers and adopting the new technologies introduced by the caGrid development team, MSI will continue to be at the forefront of caGrid software development for years to come.

For more information about MSI's caGrid projects and to download your own copy of the software, please see: https://www.msi.umn.edu/development/tropix.html