This is a test version of Biostars. For the public version, visit https://www.biostars.org.
Bioinformatics Specific I/O Benchmark?

I am trying to evaluate a particular vendors storage array and need an I/O benchmark that is bioinformatics specific. I have run the usual IT benchmark programs such as iozone, dd, bonnie++ etc but need something bio specific.

I have also run various bio codes (gromacs, NAMD, bowtie) but none of these seem to do any major I/O.

Can anyone suggest anything that would be of use? Please note that I am an IT guy not a bio guy so go easy!

Thanks, Phil.

• 2,736 views • link
updated by Biostar Biostar • written by laxton.phil • laxton.phil

2 answers

This is a very broad question. Most bioinformatics tools seem to be CPU-bound, and so I/O should not be much of an issue. If your definition of bioinformatics includes building databases, however, I think the usual database I/O benchmarks should still be informative. In general, most bioinformatics tools don't generate too many IOPs, since the most labs are not experts in high performance computing.

• 78 views • link
written by Matt Shirley Matt Shirley

I certainly think a network of small nodes will be saturated during typical NGS analyses.

Most bioinformatics setups will require a lot of configuration. Perhaps running some of the scripts in the ENCODE virtual image might be your best bet: http://scofield.bx.psu.edu/~dannon/encodevm/

• 23 views • link
written by Jeremy Leipzig Jeremy Leipzig

Log in to answer this question.

More posts like this