What Academic Hpc (Or Cloud) Infrastructures For Ngs Data Analysis And Storage Are There?

Question

[Samuel Lampa]

I'm trying to compile an overview of (academic) infrastructures out there solving the Next Generation Sequencing data analysis and storage problem by utilizing HPC (-like) resources. I'm interested mainly in academic infrastructures, and also leaning more towards HPC based solutions than cloud-based ditto, even if the latter one is interesting to take in for comparison.

If you happen to know of any such infrastructure, feel free to mention it, and if you can, provide some or all of the following information:

• Website URL.
• Any publication describing the infrastructure.
• Resource size: Approx. size of the resource (number of cpu cores, nodes, RAM per node, storage available etc).
• Software: How is the plethora of NGS software taken care of? Who installs/maintains it, etc?
• Reference genomes etc. How is that dealt with?
• User interfaces: What means of accessing the system is there? Command line only? Any graphical user interfaces (like "Galaxy")?
• User training: How is that dealt with?
• Any other comments.

Answer 1

[Istvan Albert]

I know of GenomeSpace:

http://www.genomespace.org/ funded by the National Human Genome Research Institute, with additional support from Amazon Web Services.

integrates with a number of other software tools Galaxy, IGV, Cytoscape and others.

Answer 2

[Sean Davis]

Here is a link to the NIH cluster, biowulf:

http://biowulf.nih.gov/

It is centrally maintained, but it is a general-purpose linux-based cluster, so users are free to maintain their own resources. There is a Galaxy instance running on it, but main usage is command-line. User training is done via both community-effort and through the biowulf staff.

[Samuel Lampa]

Interesting! Indeed, they seem to have quite a number of the typical NGS tools istalled: http://biowulf.nih.gov/apps/#ngs

Answer 3

[zx8754]

SGI http://www.sgi.com/company_info/newsroom/press_releases/2010/january/icr.html