A fast, scalable, MinHash-based k-mer tool to assess Sequence Read Archive next generation sequence submissions

Abstract Sequence Read Archive submissions to the National Center for Biotechnology Information often lack useful metadata, which limits the utility of these submissions. We describe a scalable k-mer based tool for fast assessment of taxonomic diversity intrinsic to submissions, independent of metadata. We show that our MinHash-based k-mer tool is accurate and scalable, offering reliable criteria for efficient selection of data for further analysis by the scientific community, at once validating submissions while also augmenting sample metadata with reliable, searchable, taxonomic terms.