Genome-wide recombination map construction from single individuals using linked-read sequencing

Abstract

Meiotic recombination is a major molecular mechanism generating genomic diversity. Recombination rates vary across the genome, often involving localized crossover \"hotspots\" and \"coldspots\". Studying the molecular basis and mechanism underlying this variation within and among individuals has been challenging due to the high cost and effort required to construct individualized genome-wide maps of recombination crossovers. In this study we introduce a new method to detect recombination crossovers across the genome from sperm DNA using Illumina sequencing of linked-read libraries produced using 10X Genomics technology. We leverage the long range information provided by the linked short reads to phase and assign haplotype states to each DNA molecule. When applied to DNA from gametes of a diploid organism, the majority of linked-read molecules can be used to faithfully reconstruct an individuals two haplotypes present at each location in the genome. A valuable rare fraction of molecules that span meiotic crossovers between the two chromosome haplotypes can then be isolated from the broader population of nonrecombinant molecules. Our pipeline, called ReMIX, allows us to characterize the genomic location and intensity of meiotic crossovers in a single individual and faithfully detects previously described recombination hotspots discovered by studies using mapping panels in mice. With a median crossover resolution of the mouse and stickleback being 15kb and 23kb respectively, ReMIX provides a powerful, high-throughput, low-cost approach to quantify recombination variation across the genome opening up numerous opportunities to study recombination variation with high genomic resolution in multiple individuals. ReMIX source code is available at at https://github.com/adreau/ReMIX.