Genome-Wide Recombination Mapping in Village Dogs

DNA sequencing of 51 village dogs was performed using Illumina technology to 8–12 fold coverage, using 101 base-pair paired end reads (Table S1 and Table S2). Reads were aligned to the reference genome using bwa[31] (link). Variant calls were made using GATK[32] (link), and phased using BEAGLE[33] (link). Extensive filters were applied to ensure that only high quality variants were used for the purposes of recombination rate estimation (see Supplementary Text S1 for details). After filtering, recombination rates were estimated using the statistical package, LDhat[15] (link).
For H3K4me3 ChIPseq experiments, spermatocytes of various stages were cell types were purified sedimentation velocity (STA-PUT) of collagenase digested single cell suspensions. Chromatin immunoprecipitation (ChIP) of H3K4me3 was performed using standard procedures, and validated through qPCR (Table S6). Libraries were prepared using Tru-Seq adaptors, with sequencing performed using 150 bp paired-end reads from an Illumina HiSeq 2500 in Rapid Run mode. Reads were mapped to the canine reference genome, and H3K4me3 peaks called using MACS[34] (link).
Detailed methods are available in the Supplementary Information. Genetic maps and called hotspots are available for download from: http://autonlab.einstein.yu.edu/dog_recomb/