Optimising Genome-Wide Detection of Runs of Homozygosity: Impacts of Reference Genome Quality and Sequencing...

Inbreeding and inbreeding depression pose a critical challenge to the persistence of small and isolated populations, driving the need for precise assessment of genomic metrics. Genome-wide runs of homozygosity (ROH) have been widely used for evaluating contemporary inbreeding levels and tracing historical events, circumventing the limitations of methods based on pedigree records. However, the reliability of ROH detection is contingent upon the quality of both the reference genome and resequencing data. Here, we employed a simulation-based approach, generating an inbred population with individuals exhibiting varying inbreeding coefficients and 13 reference genomes with differing levels of contiguity. This framework enabled us to systematically investigate the effects of sequencing depth, read length, reference genome continuity and the phylogenetic divergence of reference genomes on detecting genome-wide ROH segments. We found that a sequencing depth of ≥ 15× and a reference genome with a contig N50 > 4 Mb enabled discrimination of both the recent and historical inbreeding events, and a reference genome of congeneric subspecies is an optimal choice for ROH detection if a species-specific reference genome is not available. Furthermore, we performed parameter optimisation for PLINK to enhance ROH detection accuracy under low-coverage sequencing data and imperfect reference genomes. Our findings established methodological guidance for improving ROH-based inbreeding assessments, providing critical insights for conservation genomics and breeding programmes where accurate characterisation of genomic homozygosity is paramount.