About BioHPC @ UT Southwestern

The UT Southwestern BioHPC team provides and maintains high-performance computing, storage and client systems for the UTSW research community. BioHPC offers hardware for cluster computing, large data storage, file sharing inside and outside the campus, and a portal to integrated desktop and thin-client computing in individual labs. These offerings are now made available to users across campus, both in the basic sciences and the clinical arenas.

The BioHPC Business Plan summarizes how academic and clinical teams may interact with this infrastructure, including details about the process of gaining membership to the BioHPC. Current BioHPC membership includes:

  • Center for Alzheimer’s and Neurodegenerative Diseases
  • Cecil H. and Ida Green Center for Reproductive Biology Sciences
  • Children's Research Institute
  • Departments of Cell Biology
  • Department of Biophysics
  • Department of Immunology
  • Department of Pathology
  • Department of Radiology
  • Department of Urology
  • Department of Psychiatry
  • Department of Internal Medicine
  • Green Center for Systems Biology
  • Harold C. Simmons Comprehensive Cancer Center
  • Lyda Hill Department of Bioinformatics
  • McDermott Center for Human Growth and Development
  • Texas Institute for Brain Injury and Repair

The BioHPC infrastructure is operated by a team of experienced computational scientists under the leadership of Liqiang Wang. This development team, as well as administrative support, are co-located with the Lyda Hill Department of Bioinformatics, with hardware infrastructure housed in various spaces throughout campus. The development team covers all aspects of HPC for users, from system expansion and maintenance to carrying out collaborative work on computational research projects. Development work is focused on providing easy access for researchers to the BioHPC infrastructure. The team’s primary mission is to lower the barriers for use of HPC systems for expert and novice users alike, thus allowing more UTSW researchers to work efficiently with computationally complex analyses and big data. Day-to-day inquries are handled via our biohpc-help ticket system which is monitored by all staff.

 


 

Our Services

servers.jpg

Current BioHPC services include (as of Mar. 2018):

  • A 11520 CPU Cores HPC cluster, nucleus, coupled to fast storage systems.
  • 32 large memory GPU equipped nodes for massively parallel programs using the NVIDIA CUDA toolkit.
  • 2500TB hign-end (30-40GB/s max. throughput) Lustre filesystem (/project), and 5530TB (5.4PB) high-end GPFS parallel filesystem (/work and /archive). 
  • A wide range of installed software including compilers, MPI libraries, Matlab, R, Python, common bioinformatics software and the SBGrid distribution.
  • Provision of Linux workstations and thin-clients connected to BioHPC systems.
  • Virtual machine images for access to other operating systems.
  • Cloud-based file storage, including an external file transfer system.
  • Git repository hosting using GitLab for software version control.
  • A comprehensive web-based user portal for access to news, status, training and other resources.
  • NGS Pipeline - a customized web-based pipeline for performing common NGS workflows using the BioHPC cluster.
  • Galaxy - an installation of the Galaxy platform configured to use BioHPC systems.

storage_lights.jpg

Future intended developments:

  • Improved virtual machine images - a comprehensive collection of Windows and Linux OS images, with common software, deployable for users via the web portal.

If you have any ideas for developments that would make our systems easier, or more convenient to use please let us know via the biohpc-help email address.