The study, led by Chang Huang, Zhiyuan Ren, and Nengshui Ding of Jiangxi Agricultural University together with colleagues at Aonong Group and Jiangxi Science and Technology Normal University, focused on runs of homozygosity, or ROH. These are continuous segments of the genome in which an animal carries two identical copies of the same ancestral DNA sequence, one from each parent. When two parents share recent common ancestors, their offspring are likely to inherit matching segments on both sides, producing these long homozygous tracts. Because their length reflects how far back in the pedigree the shared ancestor lived, ROH provide a genome-wide, DNA-level measure of inbreeding that is often more accurate and more informative than traditional pedigree-based estimates, which depend on complete and accurate record keeping that commercial herds rarely have.

To capture this signal, the team analyzed genotypes from 5,859 pigs after stringent quality control, drawn from the three American breed populations. They defined ROH as segments of at least one megabase, one million DNA letters, and restricted their search to autosomes 1 through 18, excluding the sex chromosomes to keep the analysis comparable across males and females. Across all animals they detected a remarkable 416,225 individual ROH segments. That sheer number illustrates how pervasive autozygosity is in commercial breeding populations, where decades of selection for lean growth, litter size, and carcass quality have reshaped the genome through both deliberate selection and the unavoidable relatedness that accumulates in closed breeding lines.

One of the clearest findings was that the three breeds differ in their overall ROH burden. American Duroc pigs showed the highest individual-level ROH of the three populations. This pattern is consistent with the breeding structure of terminal sire lines, which are often maintained in relatively small nucleus herds under intense selection for growth and meat quality, conditions that tend to concentrate ancestral haplotypes. Landrace and Yorkshire, maternal breeds selected primarily for reproductive performance, showed different ROH architectures. The differences among breeds matter because they suggest that the historical sources of inbreeding, whether from small effective population sizes, founder effects, or selection sweeps, have acted with different intensities in each genetic line.

The heart of the study, however, lay in asking whether this autozygosity has consequences that can be measured in the barn. The researchers computed for each animal a genomic inbreeding coefficient, FROH, defined as the proportion of the autosomal genome covered by ROH. They then tested whether FROH was associated with a panel of production-related traits, using mixed statistical models that included a genomic relationship matrix. This matrix accounts for the ordinary kinship among all animals in the dataset, a crucial control, because relatives resemble each other both genetically and phenotypically for reasons that have nothing to do with inbreeding per se. Without this correction, any association between FROH and a trait could simply reflect family structure rather than the specific effect of homozygous segments.

Across 51 population-by-trait tests, the team identified 19 nominally significant associations between FROH and production traits at a probability threshold of 0.05. Thirteen of these survived correction for the false discovery rate using the Benjamini-Hochberg procedure, a statistical safeguard that limits the expected proportion of false positives among the declared findings. The pattern of surviving associations was strikingly uneven. American Landrace showed the broadest set of FDR-supported associations, Duroc had fewer, and in Yorkshire the nominal associations were restricted to body length and body height, with only the body height association holding up after correction. This breed-by-breed heterogeneity is the study’s central message: the genomic consequences of inbreeding are not a universal constant but depend on which population is being examined and which trait is being measured.

Why should different breeds show different inbreeding effects? The likely explanation lies in the concept of inbreeding load. Recessive deleterious variants are scattered across every genome, but their identity and genomic location differ among populations, shaped by each breed’s unique history of selection and drift. When autozygous segments arise, they expose whatever recessive variants happen to lie within them. If a breed’s deleterious burden happens to overlap regions influencing, say, body length or backfat, inbreeding will depress those particular traits. If the burden lies elsewhere, different traits will suffer. The population-specific associations observed here are exactly what this model predicts, and they argue strongly against a one-size-fits-all approach to managing inbreeding in multi-breed breeding programs.

The study also probed two more subtle questions. First, the researchers asked whether the effect of FROH differed between males and females, testing FROH-by-sex interactions. They found 45 estimable interaction tests, with some nominally significant results in Duroc, but none of these survived false discovery rate correction, meaning the evidence for sex-specific inbreeding effects remains tentative at best. Second, they used segmented regression models to look for thresholds, points beyond which additional autozygosity might cause a disproportionate decline in performance. Here they detected 25 nominally significant threshold responses, but again none survived FDR correction, with the smallest corrected value reaching 0.118. Importantly, the candidate breakpoints suggested by these exploratory analyses differed by population, trait, and ROH length class, hinting that if true thresholds exist, they are not universal values but properties of each population’s genome.

The distinction between ROH length classes deserves attention because it carries biological meaning. Short ROH reflect ancient shared ancestors, often reaching back many generations or even to the founding of the breed, while long ROH flag recent inbreeding, typically within the last few generations. Long ROH are therefore the ones most responsive to current mating decisions. The observation that associations and candidate thresholds varied across length classes suggests that old and new inbreeding may contribute differently to trait variation, a finding that could eventually help breeders distinguish the inbreeding they can control through mating plans from the legacy burden embedded in the breed’s history.

For the pork industry, the practical implications are concrete. Genomic inbreeding coefficients computed from routine SNP chip data could be incorporated into selection indices and mate allocation algorithms, allowing breeding companies to balance genetic gain against the accumulation of autozygosity on a population-specific basis. Rather than imposing a generic inbreeding ceiling, programs could weight the traits that are demonstrably sensitive to FROH in each line. The authors note that their findings provide a basis for population-specific evaluation of inbreeding load and for improving genomic management strategies in commercial pig breeding. As genomic selection accelerates the rate of genetic change, it also accelerates the loss of genetic diversity, making tools that measure and manage autozygosity directly from DNA increasingly essential. This study demonstrates both the promise of that approach and the statistical caution needed to interpret it: the strongest signals are real and breed-specific, while the more tantalizing patterns of sex effects and inbreeding thresholds remain, for now, hypotheses for larger datasets to test.

Subject of Research: Associations between runs of homozygosity-based genomic inbreeding and production traits in three commercial pig populations

Article Title: Runs of homozygosity reveal associations between autozygosity and production traits in three commercial pig populations

Article References: Huang, C., Ren, Z., Yang, W., Meng, Q., Wu, G., Chen, H., & Ding, N. (2026). Runs of homozygosity reveal associations between autozygosity and production traits in three commercial pig populations. BMC Genomics. https://doi.org/10.1186/s12864-026-13326-1

Image Credits: AI Generated

DOI: 10.1186/s12864-026-13326-1

Keywords: runs of homozygosity, autozygosity, genomic inbreeding, pigs, Duroc, Landrace, Yorkshire, production traits, FROH, genomic selection, animal breeding, quantitative genetics