Evaluation of SMP code in a ScaleMP environment

Project Details

Project Lead
Chris Reidy 
Project Manager
Chris Reidy 
Project Members
Chris Reidy  
Supporting Experts
 
Institution
University of Arizona, UITS  
Discipline
Biosciences, n.e.c. (617) 

Abstract

We intend to run large Trinity jobs on Echo that do not run on a conventional supercomputer due to the requirement for large addressable memory. Specifically we are testing the ability of ScaleMP to handle assembly of large volumes of RNA-seq reads. Andrew Younge (ajyounge) will be able to assist us as an expert or helper. This is anticipated to take a couple of weeks.

Intellectual Merit

The ability to process genome assembly on a large scale has value beyond the specific jobs that we intend to run. This evaluation will have wider practicability for de novo sequencing in life sciences.

Broader Impacts

Other applications like AbySS, bowtie, fastx, hmmr,soapdenovo and velvet are used by our researchers, and have a similar computational environment. If we can efficiently run these Trinity jobs, we can use these results for the other life science applications.

Scale of Use

We have two jobs to run of different size. They will be run multiple times with tuning parameters provided by ScaleMP (already arranged), to test efficacy and efficiency. The duration is likely to be two weeks