Abstract:
The declining costs of genome sequencing and growing amounts of genetic data has allowed the field of genomics to become more integrated with computational analysis. The use of high performance clusters (HPC) is necessary to compute the large amounts of data in genomic projects, however, many biologists lack background experience in working with HPC systems, which limits their ability to best address their research questions. The National Center of Genome Analysis Support (NCGAS) is an NSF-funded center that focuses on filling this need, by providing training as workshops, bioinformatics support on projects, and access to compute resources. As a byproduct of helping research projects, we develop open source workflows and make them available to the community. Here we present a developed workflow that will assist researchers in mining the Sequence Read Archive (SRA), to identify environments/datasets potentially containing genomes of interest, and identify their closely related genomes. As a proof of concept, we used two genomes to test the developed workflow, selected to ensure the flexibility of the workflow to generate results in formats amiable to further downstream analysis, based on the research question. The developed pipeline is made available through GitHub (https://github.com/NCGAS/CEWiT-REU-Identifying-datasets-in-SRA-using-Jetstream), and available as a pre-installed workflow on the XSEDE Jetstream cloud computing infrastructure.
