Runs of Homozygosity (RoH)
RoH were called based on the density of heterozygous sites in the genome using the implemented Hidden Markov Model (HMM) in bcftools roh (55) with the following parameters: -G30 –skip-indels –AF-dflt 0.4 –rec-rate 1e-9 from the mappable fraction of the genome with the filtered VCF file. We kept the RoH with a phred score > 75. We divided the RoH into different size bins: very short RoH (<100 kb), short RoH (100-500 kb), intermediate RoH (500 kb-1Mb) and long (1-5 Mb or > 5Mb). Short RoH reflect LD patterns, intermediate size RoH describe background inbreeding due to genetic drift and long RoH appear in the case of recent inbreeding (56). Results were plotted in R v 3.6.3 (57).