Using COmanage to Provision Bioinformatics Applications in High Performance Computing
Time 10/07/15 02:55PM-03:20PM
Room Room 26-A
Virtual Organizations typically provide access to web applications such as wikis, calendars, and mailing list managers. However, life science researchers are expanding their collaborations, leveraging global centers that have developed specialized skills, and building distributed computing centers with systems/applications that need to act as service providers. This session details the use of COmanage to realize the potential of trust federation-based virtual organizations for computational biology applications installed at the African Center of Excellence in Bioinformatics in Bamako, Mali. While complicated, the patterns for using SAML federated identities with web applications are well understood, but for most UNIX-based HPC applications and now some that run on Windows, shell access is the only option. The session will review the requirements needed for federated access to shell applications in comparison with options available in SAML authentication, the workarounds and methods used to provision shell access for these platforms, and specific examples in the domains of computational biology and genomics research. This will include a demonstration of the mechanisms provided by the collaboration management infrastructure (based on COmanage) and how it facilitates the bioinformatics community’s access to command-line and graphical user interfaces through the use of common technologies such as SSH and LDAP. Examples of functionality will include mixed-use service providers that host both web applications such as Galaxy (genomic sequence manipulation) as well as shell applications including R (gene expression analysis), UGENE (gene sequence analysis), and Inkscape (scientific illustration).
Speaker Matthew Economou National Institutes of Health (NIH)
Speaker Benjamin Oshrin Spherical Cow Group
Primary track Trust and Identity