Runs of Homozygosity in Population Genetics and Genomics

  • Runs of homozygosity (ROH) are continuous regions of the genome in which an individual carries homozygous genotypes at consecutive genetic markers. In diploid organisms, each chromosome is inherited as one copy from each parent. When the two chromosome copies contain the same alleles across an extended genomic region, that region may appear as a run of homozygosity. ROH are important in population genetics, human genetics, animal and plant breeding, medical genomics, and conservation biology because they provide information about shared ancestry, inbreeding, genetic diversity, and population history.
  • Homozygosity occurs when an individual carries two identical alleles at a particular genetic locus. For example, at a biallelic SNP with alleles A and G, the genotypes AA and GG are homozygous, while AG is heterozygous. A single homozygous SNP does not provide strong evidence of recent shared ancestry because common alleles may occur in unrelated individuals. However, when many consecutive SNPs are homozygous across a chromosome segment, the pattern may indicate that the individual inherited the same ancestral chromosome segment from both parents.
  • ROH are closely related to identity by descent (IBD) and autozygosity. A genomic region is autozygous when the two chromosome copies in an individual descend from the same ancestral copy. Such a region is therefore identical by descent between the two homologous chromosomes. ROH are commonly used as an observable genomic indicator of autozygosity, but the two concepts are not identical. A homozygous region may occur by chance or because of common allele frequencies and population history, without necessarily representing recent autozygosity. The likelihood that a particular ROH reflects shared ancestry depends on its length, marker density, allele frequencies, and the demographic history of the population.
  • ROH form because of several evolutionary and reproductive processes. One important cause is mating between related individuals. When parents share common ancestors, they may carry chromosome segments inherited from the same ancestral source. Their offspring can inherit matching copies of these segments from both parents, creating long homozygous regions. This process is known as inbreeding and increases the probability of autozygosity. Repeated mating among relatives across generations can increase the number and total length of ROH in a population.
  • Population bottlenecks and founder effects can also increase ROH. A bottleneck occurs when a population experiences a substantial reduction in size, resulting in the loss of genetic diversity and increased genetic drift. A founder effect occurs when a new population is established by a relatively small number of individuals from a larger population. In both cases, descendants may inherit chromosome segments from a limited ancestral gene pool. This can increase background homozygosity and the frequency of shared genomic segments, even when close-relative mating has not occurred in every generation.
  • Small effective population size is another important factor. Effective population size represents the size of an idealized population that would experience the same rate of genetic drift or inbreeding as the population under study. Populations with small effective sizes generally experience stronger genetic drift and more rapid accumulation of relatedness. Over generations, this can increase the frequency and extent of ROH. Consequently, ROH patterns can provide useful evidence about long-term population size and demographic history.
  • Recombination influences the length of ROH inherited from common ancestors. During meiosis, homologous chromosomes exchange genetic material, breaking ancestral chromosome segments into smaller pieces. When the common ancestor is relatively recent, fewer generations of recombination have occurred, so the inherited shared segments are more likely to remain long. When the common ancestor lived many generations ago, repeated recombination generally produces shorter segments. Thus, long ROH often indicate recent parental relatedness, whereas shorter ROH may reflect more distant shared ancestry or historical reductions in population size. This relationship is probabilistic rather than exact because recombination rates vary across the genome and among genomic regions.
  • ROH are commonly classified into length categories to investigate different timescales of ancestry. Long ROH are often associated with recent parental relatedness, while shorter ROH may reflect older demographic events, background relatedness, or population-specific linkage disequilibrium. Some studies use categories such as 1–2 megabases, 2–4 megabases, 4–8 megabases, and greater than 8 megabases, but these boundaries are not universal. The appropriate thresholds depend on the species, genome size, marker density, recombination rate, population history, and research objective. A length category should therefore be interpreted in the context of the dataset rather than treated as a fixed estimate of the number of generations since a common ancestor.
  • The total length of ROH in an individual is often used to estimate genomic inbreeding. One widely used measure is the proportion of the autosomal genome covered by ROH. It is commonly represented by the coefficient (F_{\mathrm{ROH}}), calculated as the total length of detected ROH divided by the total length of the autosomal genome considered in the analysis.

[
F_{\mathrm{ROH}} =
\frac{\text{Total length of ROH}}{\text{Total autosomal genome length}}
]

  • For example, if an individual has 200 megabases of detected ROH across an autosomal genome of 2,500 megabases, the estimated (F_{\mathrm{ROH}}) is 0.08, or 8%. This means that 8% of the genomic length included in the analysis is covered by detected ROH. The estimate depends on the ROH detection criteria, marker coverage, and genome length used as the denominator. It should not be interpreted as a universal measure of pedigree inbreeding without considering these methodological choices.
  • Different ROH-based measures can provide additional information. The number of ROH in an individual describes how many separate homozygous segments have been detected. The average ROH length provides information about segment-size distribution, while the total ROH length measures the overall genomic extent of homozygosity. Researchers may also calculate the proportion of the genome covered by ROH in different length categories. Comparing these measures can help distinguish patterns associated with recent parental relatedness from those associated with older population history.
  • The distribution of ROH across individuals can reveal differences in inbreeding and genetic diversity within a population. Individuals with many long ROH may have parents who share relatively recent ancestors, while individuals with numerous shorter ROH may reflect older shared ancestry or a historically small population. However, these patterns are not definitive on their own. Population structure, recombination-rate variation, marker ascertainment, and technical quality can influence ROH length and frequency. Analyses should therefore consider pedigree records, population structure, and other genomic measures where available.
  • ROH are also useful for identifying genomic regions that are frequently homozygous across individuals in a population. Such regions are sometimes called ROH islands or commonly shared ROH regions. They may occur because of historical selection, genetic drift, population bottlenecks, or other demographic processes. In some cases, a region may contain genes associated with important biological traits or disease susceptibility. However, a frequently observed ROH region is not automatically evidence of positive selection. Demonstrating selection requires additional analyses, appropriate population comparisons, and consideration of alternative demographic explanations.
  • In human genetics, ROH analysis is used to investigate consanguinity, recessive disease risk, population history, and genomic diversity. Individuals whose parents are related may carry long autozygous regions, increasing the probability that rare recessive variants are present in homozygous form. ROH can therefore help prioritize genomic regions in which disease-causing variants may be located, particularly in families affected by suspected recessive disorders. In clinical and research settings, ROH findings must be interpreted alongside clinical features, family history, sequencing results, and variant evidence. A homozygous region alone does not identify a disease-causing gene or establish a diagnosis.
  • In animal breeding, ROH are important for monitoring genomic inbreeding and managing breeding populations. Livestock populations may experience increased homozygosity when a small number of popular breeding males contribute disproportionately to subsequent generations. The resulting reduction in effective population size can increase the accumulation of relatedness and ROH. Genomic ROH analysis can help breeders identify animals with high levels of autozygosity, monitor changes in inbreeding across generations, and design mating strategies that reduce the risk of excessive inbreeding.
  • ROH analysis can also support the study of economically important traits in livestock and plants. Long homozygous regions may contain recessive alleles affecting fertility, survival, growth, production, disease susceptibility, or other traits. In some breeding populations, selection for desirable characteristics may increase homozygosity around particular genomic regions. However, the presence of an ROH near a candidate gene does not prove that the region has been selected or that a particular allele is responsible for a trait. Association studies, functional evidence, pedigree information, and comparisons with suitable control populations are needed to investigate these possibilities.
  • In conservation genetics, ROH provide information about genetic erosion, population isolation, and the consequences of small population size. Endangered populations may contain elevated levels of homozygosity because of bottlenecks, restricted gene flow, and mating among relatives. Monitoring ROH can help conservation researchers assess genomic inbreeding and identify populations that may be at risk of reduced genetic diversity. Such information can contribute to decisions about population management, translocation, and genetic rescue. These decisions should also consider local adaptation, disease risks, ecological conditions, and the potential effects of introducing individuals from other populations.
  • ROH are detected using genome-wide SNP genotyping arrays or whole-genome sequencing data. SNP arrays provide genotypes at selected marker positions and are often used in large population studies because they can be relatively cost-effective. Whole-genome sequencing provides much denser information and can improve the detection of smaller ROH, although results depend on sequencing depth, variant calling quality, and filtering. The data are generally organized by chromosome, genomic position, and genotype, allowing software to identify consecutive homozygous markers.
  • ROH detection algorithms examine genotype patterns across chromosomes and apply criteria to define a continuous homozygous region. Common parameters include the minimum number of consecutive homozygous SNPs, minimum physical length, maximum number of missing genotypes, maximum number of heterozygous calls, and permitted gap between consecutive markers. Some methods also consider marker density and the frequency of heterozygous genotypes expected by chance. These parameters strongly influence the number and length of detected ROH, so they should be selected carefully and reported clearly.
  • PLINK is one of the widely used tools for ROH analysis. It can identify homozygous segments from genome-wide genotype data using user-defined criteria. The following PLINK 1.9 example illustrates a basic ROH analysis:

plink –bfile genotype_data \

      –homozyg \

      –homozyg-snp 50 \

      –homozyg-kb 1000 \

      –out roh_analysis

  • In this example, –bfile specifies the input PLINK binary dataset, –homozyg activates ROH detection, –homozyg-snp 50 requires at least 50 consecutive SNPs, and –homozyg-kb 1000 sets a minimum segment length of 1,000 kilobases. These are illustrative settings, not universal recommendations. The appropriate values depend on the organism, marker density, genome characteristics, and research question. Users should check the documentation for their installed PLINK version because options and defaults may differ.
  • PLINK output can be used to calculate the total length of ROH for each individual and estimate (F_{\mathrm{ROH}}). The following Python example illustrates this calculation from a table containing individual identifiers and ROH segment lengths in kilobases. It assumes that the table contains one row per detected ROH and that the genome length is supplied in kilobases.

import pandas as pd

# Example ROH data

roh_data = {

    “individual”: [“Animal_1”, “Animal_1”, “Animal_2”, “Animal_2”],

    “roh_length_kb”: [2500, 1800, 1200, 900]

}

df = pd.DataFrame(roh_data)

# Total ROH length per individual

total_roh = (

    df.groupby(“individual”)[“roh_length_kb”]

      .sum()

      .reset_index(name=”total_roh_kb”)

)

# Example autosomal genome length in kilobases

genome_length_kb = 2_500_000

# Calculate F_ROH

total_roh[“F_ROH”] = (

    total_roh[“total_roh_kb”] / genome_length_kb

)

print(total_roh)

  • The genome length used in this calculation should correspond to the autosomal regions included in the ROH analysis. If only selected chromosomes or genomic intervals were analysed, the denominator should reflect the analysed regions rather than the entire genome. Sex chromosomes and regions with poor marker coverage may require separate treatment.
  • Other tools can also support ROH analysis. BCFtools and VCFtools are useful for processing and filtering variant data before downstream analysis. R packages such as detectRUNS provide functions for identifying and summarizing ROH from genotype data. SNPRelate supports large-scale genotype analysis and related genomic calculations. Specialized tools and pipelines may be used for sequence-based ROH detection or for particular species and data types. Researchers should examine the assumptions and parameter settings of each method before comparing results across studies.
  • Quality control is essential before ROH detection. Genotype errors can break a true homozygous segment or create an apparent homozygous region. Missing genotypes, low-quality variants, incorrect chromosome positions, and uneven marker density can also affect segment detection. Very low-density marker datasets may fail to detect short ROH, while some marker arrays can introduce ascertainment bias because their SNPs were selected using particular populations. Whole-genome sequencing reduces some marker-selection limitations but can still be affected by coverage variation, mapping errors, and variant-calling errors.
  • The interpretation of ROH also requires attention to linkage disequilibrium (LD). Nearby markers are often correlated because of limited recombination, and this can create extended regions of similar allele patterns. Long stretches of homozygosity may therefore reflect a combination of autozygosity and population-specific haplotype structure. The distinction is particularly important when comparing populations with different demographic histories or marker panels. Using appropriate population references and consistent analytical settings helps reduce misleading comparisons.
  • ROH should not be treated as a direct substitute for pedigree-based inbreeding coefficients. Pedigree estimates depend on the completeness and accuracy of recorded ancestry and describe expected inbreeding under the pedigree model. (F_{\mathrm{ROH}}) measures the proportion of the analysed genome covered by detected homozygous segments and reflects realized genomic patterns. The two measures may differ because of random inheritance, recombination, pedigree depth, missing ancestors, and technical properties of the genomic data. Combining pedigree and genomic information can provide a more complete assessment of inbreeding.
  • For reproducible ROH studies, researchers should report the reference genome assembly, genotype data source, quality-control procedures, marker density, software and version, ROH detection algorithm, and all thresholds used. Important parameters include minimum segment length, minimum SNP count, permitted heterozygous calls, missing genotype allowance, and maximum inter-marker distance. Studies should also explain how total ROH length and (F_{\mathrm{ROH}}) were calculated and whether sex chromosomes or poorly covered regions were excluded. Transparent reporting is necessary for meaningful comparisons across populations and research projects.
  • Runs of homozygosity are valuable genomic indicators of autozygosity, shared ancestry, and population history. Their number, length, and genomic distribution can help researchers investigate recent parental relatedness, historical bottlenecks, effective population size, recessive disease risk, and genetic diversity. They are widely applied in human genetics, animal and plant breeding, and conservation biology. Accurate interpretation requires careful genotype quality control, appropriate detection thresholds, and consideration of population structure and demographic history. When combined with pedigree records, identity-by-descent analysis, and other genomic methods, ROH provide an important tool for understanding genetic relatedness and managing genetic diversity.
Author:

Leave a Reply

Your email address will not be published. Required fields are marked *