Repeat-aware detection of structural variants in the pig genome: Distinguishing true rearrangements from false positives with Oxford Nanopore Sequencing

Structural variants (SVs) are a major source of genomic diversity but also a leading cause of reduced fertility in pigs. Reciprocal translocations (RTs) disrupt chromosome pairing during meiosis, producing unbalanced gametes, embryonic lethality, and smaller litter sizes. Traditional cytogenetic methods such as karyotyping and fluorescence in situ hybridisation (FISH) can detect RTs but have limited resolution, particularly in repetitive regions where many breakpoints occur. Long-read sequencing with Oxford Nanopore Technologies (ONT) offers base-pair resolution of these events but remains challenging due to the high repeat content of mammalian genomes. To address this, we sequenced three domestic pigs (Sus scrofa), including two with RTs previously validated using FISH and Hi-C and one wild type control. A fourth individual was used for technical protocol optimisation. We optimised DNA extraction and sequencing to maximise read length and coverage, and benchmarked Sniffles2 and BigClipper against curated RT sets. Sniffles2 produced conservative calls, while BigClipper identified more unique breakends (BNDs) but with more candidate breakends in repeat-rich regions, consistent with an elevated false-positive burden. Repeat annotation with RepeatMasker, combined with enrichment analyses using regioneR and Genome Association Tester (GAT), revealed strong biases in false-positive BNDs towards long terminal repeats (LTRs) and satellites. By contrast, validated RT breakpoints were consistently flanked by short interspersed nuclear elements (SINEs), including MIR and tRNA/Pre0_SS subtypes. Read-based visualisation with Ribbon, SVbyEye, and custom LASTZ-based dotplots confirmed that these repeats contributed short tracts of homology at breakpoint junctions, consistent with repeat-mediated rearrangement. These findings support the integrative breakage model, in which rearrangements reflect the combined influence of sequence composition, chromatin organisation, and repair kinetics. By linking repeat biology, nuclear architecture, and SV calling performance, our analysis advances a repeat-aware framework for interpreting SVs. This has direct value for fertility screening in pigs, where undetected RTs reduce litter sizes and propagate through breeding populations and may also inform long-read approaches for diagnosing chromosome abnormalities in human male infertility.