- Identity by descent (IBD) and identity by state (IBS) are fundamental concepts in population genetics, statistical genetics, genomics, and genetic relatedness analysis. Both describe similarities between alleles observed in different individuals, but they differ in the evolutionary history underlying those similarities. Identity by state refers to alleles that are the same in form, whereas identity by descent refers to alleles that have been inherited from the same ancestral copy. Understanding this distinction is essential for estimating genetic relationships, studying population structure, detecting shared ancestry, analysing inbreeding, and interpreting genomic data.
- Identity by state occurs when two or more individuals carry alleles that are identical in their observed form at a particular genetic locus. For example, if two individuals both carry the nucleotide allele A at a single-nucleotide polymorphism (SNP), their alleles are identical by state at that position. IBS describes the observed similarity of alleles, but it does not establish whether the alleles originated from the same ancestral chromosome. The same allele may have arisen independently through mutation or may have become common in a population through historical processes. Consequently, individuals who share an allele by state are not necessarily closely related.
- Identity by descent refers to alleles that are copies of the same ancestral allele inherited through a common ancestor without an intervening mutation that changes the allele. For example, if two siblings inherit the same copy of a chromosome segment from their father, that segment is identical by descent between them. IBD therefore provides information about shared ancestry rather than simply matching allele states. The longer the shared IBD segment, the more recent the common ancestor is generally expected to be, although recombination, mutation, demographic history, and the number of generations involved influence segment length.
- The key difference between IBD and IBS is the distinction between ancestral origin and observed allele similarity. Two individuals may carry the same allele because they inherited it from a common ancestor, making it IBD, or because the same allele is present independently in their lineages, making it IBS without necessarily being IBD. An IBD allele is generally also IBS at the observed locus, assuming no mutation or genotyping error, but an IBS allele is not necessarily IBD. This distinction is particularly important when analysing populations in which common alleles may be shared by many unrelated individuals.
- A simple example illustrates the difference. Suppose two individuals both carry the allele G at a SNP. If both inherited that allele from the same ancestral chromosome, the allele is identical by descent. If the allele G is common in the population and the individuals inherited it through different ancestral lineages, the alleles may be identical by state without being identical by descent. At a single locus, it is usually difficult to determine the ancestral origin of an allele from its nucleotide state alone. IBD inference therefore relies on patterns across multiple linked markers, haplotypes, pedigree information, population allele frequencies, and statistical models.
- In diploid organisms, each individual carries two homologous chromosome copies, one inherited from each parent. When two individuals are compared, they may share zero, one, or two alleles identical by descent at a particular autosomal locus. These possibilities are commonly represented by the IBD coefficients Z0, Z1, and Z2. Z0 is the probability that the individuals share no alleles IBD at a locus, Z1 is the probability that they share one allele IBD, and Z2 is the probability that they share both alleles IBD. The three probabilities sum to one.
- For unrelated individuals in a large, randomly mating population, the expected IBD sharing is generally close to zero under a simple pedigree model, although distant shared ancestry may still produce detectable genomic segments. For parent and offspring, one allele at each autosomal locus is expected to be shared IBD, giving Z1 approximately equal to one. For full siblings, the expected probabilities are Z0 = 0.25, Z1 = 0.50, and Z2 = 0.25. For half-siblings, the expected probabilities are Z0 = 0.50, Z1 = 0.50, and Z2 = 0. These are theoretical expectations under standard assumptions; realized sharing across a finite genome varies because of recombination and the random transmission of parental chromosomes.
- IBD sharing is closely related to the concept of the kinship coefficient, also called the coefficient of coancestry. The kinship coefficient between two individuals is the probability that one randomly selected allele from each individual at the same autosomal locus is identical by descent. It is commonly denoted by φ. The inbreeding coefficient of an individual is the probability that the two alleles at a locus are IBD. When the parents of an individual are related, the offspring may inherit two alleles that descend from the same ancestral copy, increasing the probability of autozygosity. Under standard pedigree definitions, the inbreeding coefficient of an offspring equals the kinship coefficient between its parents.
- The distinction between IBD and IBS is also important in the study of inbreeding. Inbreeding increases the probability that an individual carries two alleles inherited from a common ancestor. Such alleles are IBD and form regions of the genome known as autozygous regions. However, homozygosity alone does not prove that two alleles are IBD. An individual may be homozygous because both parents carry a common allele, even when the copies do not originate from the same recent ancestral chromosome. This is one reason why genomic inbreeding analysis uses long homozygous segments and haplotype information rather than relying only on the number of homozygous SNP genotypes.
- Runs of homozygosity (ROH) are continuous genomic regions in which an individual carries homozygous genotypes across multiple adjacent markers. Long ROH often indicate inheritance of chromosome segments from relatively recent common ancestors, while shorter ROH may reflect older shared ancestry or background linkage disequilibrium. ROH are therefore widely used as indicators of autozygosity and genomic inbreeding. Nevertheless, ROH detection depends on marker density, genotype quality, missing data, linkage disequilibrium, and the thresholds used to define a run. A homozygous segment should not automatically be interpreted as a recent IBD segment without considering these factors.
- IBD segments are stretches of DNA inherited from a shared ancestral chromosome. During meiosis, recombination breaks parental chromosomes into smaller segments, and these segments are transmitted to descendants. As generations pass, shared ancestral segments generally become shorter because of recombination. Consequently, long IBD segments often indicate relatively recent common ancestry, whereas shorter segments may reflect more distant relationships. The relationship between segment length and the number of generations to a common ancestor is probabilistic rather than exact, because recombination occurs at variable positions and rates across the genome.
- IBD segment detection is commonly performed using dense SNP genotyping data or whole-genome sequencing. Methods identify regions in which two individuals share compatible haplotypes over continuous genomic intervals. Some approaches use phased haplotypes, while others infer sharing directly from genotype patterns. Algorithms such as GERMLINE, Refined IBD, and hap-IBD identify shared genomic segments using different strategies. The choice of method depends on marker density, sample size, sequencing or genotyping quality, phasing accuracy, and the expected degree of relatedness.
- IBS measures can be calculated directly from observed genotypes. At a biallelic SNP, two individuals may share zero, one, or two alleles by state. For example, individuals with genotypes AA and AA share two alleles IBS; individuals with AA and AG share one allele IBS; and individuals with AA and GG share no alleles IBS. These comparisons can be summarized across many SNPs to estimate genome-wide genetic similarity. However, simple IBS measures treat allele matches similarly regardless of how common or rare the alleles are. A match at a rare allele may provide more evidence of shared ancestry than a match at a very common allele.
- Allele frequencies therefore play an important role in distinguishing meaningful genetic sharing from coincidental similarity. If an allele is common in a population, unrelated individuals have a relatively high probability of carrying it. If an allele is rare, sharing that allele may provide stronger evidence of common ancestry, particularly when the same rare allele occurs within a longer matching haplotype. Statistical IBD methods use allele frequencies, linkage disequilibrium, haplotype patterns, and population models to estimate the probability that shared alleles descend from a common ancestral copy. These estimates are affected by the reference population and the accuracy of allele-frequency estimates.
- IBD and IBS are widely used in kinship and relationship inference. Parent–offspring pairs, full siblings, half-siblings, cousins, and more distant relatives have different expected patterns of IBD sharing. Genomic relationship estimation uses these patterns to infer relationships even when pedigree records are incomplete or unavailable. For example, parent–offspring pairs are expected to share one allele IBD at nearly every autosomal locus, while full siblings can share zero, one, or two alleles IBD in different genomic regions. These differences allow researchers to distinguish many classes of biological relationships using genome-wide marker data.
- In population genetics, IBD sharing provides information about demographic history, population connectivity, effective population size, and recent common ancestry. Populations that have experienced bottlenecks, founder events, or prolonged isolation may contain elevated levels of shared IBD segments. A high frequency of long shared segments can indicate recent common ancestry or recent population contraction, while patterns of shorter segments may reflect older demographic events. These interpretations require appropriate demographic models because migration, population structure, admixture, and sampling can also influence observed IBD patterns.
- IBD analysis is also used to estimate effective population size. In general, smaller effective population sizes increase genetic drift and the probability that individuals share ancestral chromosome segments. The distribution of IBD segment lengths can provide information about changes in effective population size across generations. Methods such as IBDNe use the lengths and frequencies of shared IBD segments to infer recent demographic trends. However, estimates depend on accurate segment detection, suitable recombination maps, representative sampling, and assumptions about population structure and demographic history.
- In genome-wide association studies, IBD and IBS information can help identify close relatives, detect duplicate or related samples, and account for relatedness in association models. If related individuals are treated as independent observations, statistical tests may produce inflated significance levels because their genotypes and phenotypes are not independent. Genetic relationship matrices and mixed models can account for relatedness and population structure. IBS-based genomic relationship matrices are commonly used in genome-wide analyses, while IBD-based estimates can provide more direct information about recent shared ancestry.
- IBD is particularly useful in rare disease genetics and gene mapping. Individuals affected by the same recessive disorder may share a homozygous region inherited from a common ancestor, especially in consanguineous families or populations with founder effects. Such regions can help prioritize candidate genes and disease-associated variants. In family-based studies, shared IBD segments may also help identify genomic regions that co-segregate with a trait. Nevertheless, shared ancestry alone does not establish that a particular variant causes a disease. Functional evidence, segregation analysis, population frequency, and clinical interpretation are also required.
- In animal and plant breeding, IBD and IBS are important for estimating genetic relationships, controlling inbreeding, managing genetic diversity, and improving selection decisions. Pedigree-based relationships estimate expected sharing from recorded ancestry, whereas genomic methods use marker data to estimate realized relationships. Two full siblings may have different realized genomic relationships because of random chromosome transmission and recombination. Genomic relationship estimates can therefore improve the identification of related individuals and support mate allocation, genomic selection, and the management of breeding populations. These methods are also useful in conservation genetics, where maintaining genetic diversity and avoiding excessive mating among close relatives are important objectives.
- Several software tools support IBD and IBS analysis. PLINK and PLINK 2 provide functions for estimating pairwise genetic similarity and identifying related individuals using genome-wide genotype data. KING is widely used for estimating kinship and detecting relationships, including in datasets with population structure. Beagle provides haplotype phasing and IBD-related analysis capabilities. GERMLINE, Refined IBD, and hap-IBD are used for detecting shared genomic segments. BCFtools and VCFtools support genotype data processing and filtering, while R packages such as SNPRelate and related statistical tools support genomic relationship and population analyses. The appropriate tool depends on the research objective, data type, marker density, sample size, and desired level of resolution.
- The following Python example illustrates a simple IBS comparison between two diploid individuals using genotype dosage values. The values 0, 1, and 2 represent the number of copies of the alternate allele at each SNP. This simplified example calculates the number of shared alleles by state at each locus. It does not infer identity by descent.
import numpy as np
# Genotypes: 0 = AA, 1 = AG, 2 = GG
individual_1 = np.array([0, 1, 2, 1, 0])
individual_2 = np.array([0, 2, 2, 0, 1])
# Number of alleles shared by state at each SNP
ibs_per_locus = 2 – np.abs(individual_1 – individual_2)
print(“IBS alleles per locus:”, ibs_per_locus)
print(“Total IBS alleles:”, ibs_per_locus.sum())
print(“Mean IBS alleles per locus:”, ibs_per_locus.mean())
- A simplified calculation of the proportion of alleles shared by state can be expressed as the total number of matching alleles divided by the total number of alleles compared. More advanced IBS estimators may account for allele frequencies, missing genotypes, linkage disequilibrium, and population structure. The simple dosage calculation above is useful for understanding the concept but should not be treated as a complete kinship or IBD estimator.
- PLINK can be used to estimate pairwise genome-wide identity statistics from genotype data. For example, a basic PLINK 1.9 command is:
plink –bfile genotype_data –genome –out pairwise_relationships
- The resulting output includes pairwise identity estimates such as Z0, Z1, Z2, and PI_HAT, which summarize estimated sharing under the method’s assumptions. PLINK 2 has different commands and options for relatedness analysis, so users should consult the documentation for the installed version and selected estimator. Quality control, variant filtering, and population-appropriate allele-frequency estimation should be performed before interpreting the results.
- A conceptual IBD example can be represented by a pair of haplotypes shared between two individuals. If both individuals carry the same ancestral chromosome segment inherited from a common ancestor, that segment is IBD. If they carry the same allele states at individual SNPs but the alleles originate from different ancestral copies, the matches are IBS without necessarily being IBD. Phasing helps reconstruct which alleles occur together on each chromosome, making it easier to identify extended shared haplotypes. However, phasing errors and uncertain haplotype boundaries can affect IBD segment detection.
- The interpretation of IBD and IBS estimates requires several precautions. Genotyping errors, missing data, low marker density, allele-frequency differences, linkage disequilibrium, and incorrect sample labels can produce misleading estimates. Population structure may cause individuals from the same subpopulation to share more alleles by state than individuals from different subpopulations, even when they are not close relatives. In addition, different software packages use different algorithms, thresholds, and statistical assumptions. Results from different methods should therefore not be treated as directly interchangeable without examining their definitions and parameters.
- IBD and IBS are related but distinct measures of genetic similarity. IBS describes whether alleles have the same observed state, whereas IBD describes whether alleles descend from the same ancestral copy. IBS can be measured directly from genotype data, while IBD generally requires inference using shared haplotypes, pedigree information, allele frequencies, and statistical models. Their combined use supports kinship estimation, inbreeding assessment, population history reconstruction, disease gene mapping, genome-wide association studies, and genetic management in breeding and conservation. Understanding the difference between these concepts is essential for interpreting genomic similarity accurately and distinguishing simple allele matching from evidence of shared ancestry.