Rapid, Paralog-Sensitive CNV Analysis of 2457 Human Genomes Using QuicK-mer2

Genes (Basel). 2020 Jan 29;11(2):141. doi: 10.3390/genes11020141.

Abstract

Gene duplication is a major mechanism for the evolution of gene novelty, and copy-number variation makes a major contribution to inter-individual genetic diversity. However, most approaches for studying copy-number variation rely upon uniquely mapping reads to a genome reference and are unable to distinguish among duplicated sequences. Specialized approaches to interrogate specific paralogs are comparatively slow and have a high degree of computational complexity, limiting their effective application to emerging population-scale data sets. We present QuicK-mer2, a self-contained, mapping-free approach that enables the rapid construction of paralog-specific copy-number maps from short-read sequence data. This approach is based on the tabulation of unique k-mer sequences from short-read data sets, and is able to analyze a 20X coverage human genome in approximately 20 min. We applied our approach to newly released sequence data from the 1000 Genomes Project, constructed paralog-specific copy-number maps from 2457 unrelated individuals, and uncovered copy-number variation of paralogous genes. We identify nine genes where none of the analyzed samples have a copy number of two, 92 genes where the majority of samples have a copy number other than two, and describe rare copy number variation effecting multiple genes at the APOBEC3 locus.

Keywords: copy-number variation; gene duplication; k-mer.

Publication types

  • Research Support, N.I.H., Extramural

MeSH terms

  • Algorithms
  • Computational Biology / methods*
  • DNA Copy Number Variations*
  • Evolution, Molecular
  • Gene Duplication
  • Genome, Human
  • Humans
  • Sequence Analysis, DNA / methods*