Abstract
Background
Single nucleotide polymorphism-based chromosomal microarray analysis (CMA) can detect regions of homozygosity (ROHs), which may be associated with medical conditions; however, limited ROH data, especially in East Asians, complicates clinical interpretations. We characterized ROH distributions and frequencies in a Korean population using CMA, highlighting clinically relevant findings, including suspected uniparental disomy (UPD), using standardized criteria.
Methods
We analyzed ROHs in 1,731 individuals who underwent postnatal CMA at a Korean medical center. ROHs ≥3 Mb long were detected using the CytoScan Dx platform and Chromosome Analysis Suite Dx. Suspected UPD and consanguinity were assessed per the American College of Medical Genetics and Genomics technical standards.
Results
We identified 3,962 ROHs, with 76.7% of patients carrying at least one. Common “hotspot” regions included 3p21.31p21.1 (20.3%), 11p11.2 (18.2%), 1q21.1q21.3 (17.7%), and 1p33p32.3 (12.0%). Almost all ROHs observed in >1% of patients had a median size of <5 Mb. ROH frequencies correlated negatively with chromosomal recombination rates and positively with gene densities. Additionally, 1.2% (N=21) of patients exhibited ROH patterns suggestive of UPD or consanguinity (13 suspected UPDs on imprinted chromosomes, 6 on non-imprinted chromosomes, and 2 consanguinities); 8 of 13 patients with suspected UPD were diagnosed as having imprinting disorders, with no pathogenic copy number variations detected.
Conclusions
Our population-specific ROH data for Koreans improve clinical interpretations by minimizing the risk of overinterpreting benign variants and highlight the value of standardized criteria for reliably detecting UPD and consanguinity and integrating ROH analysis into routine CMA interpretations.
Keywords: Chromosomal microarray analysis, Consanguinity, Genetic recombination, Genomic imprinting, Loss of heterozygosity, Uniparental disomy
INTRODUCTION
Chromosomal microarray analysis (CMA) has served as a first-tier test for patients with developmental delays, intellectual disabilities, autism spectrum disorders, or multiple congenital anomalies [1]. In addition to detecting copy number variations (CNVs), single nucleotide polymorphism (SNP)-based CMA can identify regions of homozygosity (ROHs) [2], which are continuous DNA stretches with identical alleles on both chromosomal copies. ROHs predominantly arise through inheritance of the same ancestral haplotype from both parents (identity by descent), and small benign ROHs are commonly observed across diverse human populations [3]. However, larger ROHs are associated with a higher risk of autosomal recessive disorders owing to the increased likelihood of homozygous pathogenic variants, a concern particularly relevant in consanguineous families [4]. Sufficiently large ROHs in a single chromosome may indicate uniparental disomy (UPD)–the inheritance of two identical copies of a chromosome from a single parent–an event associated with imprinting disorders [5–7].
ROHs have been implicated in potentially pathogenic or disease-modifying mechanisms; therefore, distinguishing benign variants from those warranting clinical attention is critical. Additionally, standardized reporting criteria are necessary to ensure consistent variant interpretation and improved genetic counseling in different clinical settings. Accordingly, universal criteria for interpreting ROHs–especially in the context of suspected UPD and consanguinity–have been proposed [4–6, 8, 9]. Building on this prior research, the American College of Medical Genetics and Genomics (ACMG) has established technical standards that classify ROHs based on their size and distribution to identify suspected cases of UPD or consanguinity [10].
Although certain ROH “hotspots” have been documented across diverse populations [11, 12], the frequency and extent of these shared homozygous segments can vary substantially among different ethnic groups [13, 14], underscoring the need for population-specific data. However, few large-scale datasets pertain to the landscape of ROHs in East Asian populations, particularly in clinical settings. Establishing baseline ROH patterns in a Korean cohort would help refine diagnostic thresholds and avoid overinterpreting benign ROHs. Determining the frequency of more unusual events in the Korean population (such as UPD and consanguinity) using universal ROH interpretation criteria is key for accurate detection. We retrospectively analyzed ROHs in 1,731 individuals who underwent postnatal CMA at a single medical center in Korea. Our objectives were to comprehensively characterize the genome-wide distribution and frequency of ROHs and to identify suspected cases of UPD and consanguinity using standardized ROH interpretation criteria.
MATERIALS AND METHODS
Study population
We retrospectively analyzed ROHs in 1,731 consecutive patients who underwent postnatal CMA for clinical indications at the Medical Genetics Center (Asan Medical Center, Korea) between 2019 and 2024, including developmental delay, intellectual disability, autism spectrum disorder, or multiple congenital anomalies. Although the patients did not formally declare their ancestry, their Korean resident registration numbers indicated that most were of Korean descent, with only a small minority likely from other ethnic backgrounds. The male: female ratio was 1.25, and the mean age was 6.5 yrs (±10.7 yrs).
ROH detection
CMA was conducted using the CytoScan Dx array platform and Chromosome Analysis Suite (ChAS) Dx software (Thermo Fisher Scientific, Waltham, USA). Autosomal ROHs were identified that were ≥3 megabases (Mb) long (the detection threshold of the platform). ROHs resulting from single-copy deletions were excluded from our analysis. ROHs in the 16p11.2p11.1 region were also excluded, as they were deemed technical artifacts after inspecting the distribution of SNP probe patterns in these regions.
Associations with the recombination rate and gene density
Known recombination rates for each human autosome [15] were used for regression analysis to examine the relationship between ROH counts and recombination rates. Gene densities were estimated by counting the total number of genes within each ROH (as identified by ChAS Dx software) and normalization according to the ROH size.
Detection of suspected uniparental disomy and consanguinity
Suspected UPD and consanguinity were identified based on the ACMG’s technical standards for interpreting large ROHs [10]. Specifically, UPD was suspected in any of the following scenarios: (i) a single interstitial ROH of ≥10 Mb on an imprinted chromosome (6, 7, 11, 14, 15, or 20), (ii) a single interstitial ROH of ≥15 Mb on a non-imprinted chromosome, (iii) multiple interstitial ROHs with a total size of ≥15 Mb, or (iv) a single telomeric ROH of ≥5 Mb. Consanguinity was suspected when ROHs spanning multiple chromosomes comprised ≥3% of the total autosomal genome, corresponding to approximately a fourth-degree relationship.
Analysis of concurrent CNVs
The pathogenicity of all CNVs identified using CMA was assessed following the ACMG’s guidelines for interpreting and reporting constitutional CNVs and was classified as benign, likely benign, of uncertain significance, likely pathogenic, or pathogenic [16]. Two certified medical genetics professionals reviewed each patient’s clinical and phenotypic information to ensure accurate CNV interpretations.
Statistical analysis
Statistical analyses were performed using R software (version 4.1.3; R Core Team, R Foundation for Statistical Computing, Vienna, Austria).
Ethical considerations
This study was conducted in accordance with the Declaration of Helsinki and was approved by the Institutional Review Board of Asan Medical Center, Seoul, Korea (approval number 2024-2341), which granted a waiver of informed consent owing to the retrospective study design and reliance on reviewing medical records.
RESULTS
Genome-wide landscape of ROHs
To characterize the genome-wide distribution and frequency of ROHs, we analyzed chromosomal microarray data for 1,731 postnatal patients, identifying 3,962 ROHs, with 76.7% of patients carrying at least one ROH. We first examined the number and size of ROHs per individual. The number of ROHs per individual exhibited a right-skewed distribution with a median of value of 1 (Fig. 1A). Similarly, the ROH size also showed a right-skewed distribution, with a median length of 3.5 Mb (mean±SD: 4.4±7.7 Mb; Fig. 1B).
Fig. 1. Distribution of the numbers and sizes of regions of homozygosity (ROHs). (A) Frequency distribution of the number of ROHs per patient, showing a right-skewed pattern with a median value of 1. (B) Frequency distribution of ROH sizes. As shown in the inset, the ROH sizes were right-skewed, with a median length of 3.5 megabases (Mb). Because of the 3 Mb resolution threshold of the CytoScan HD platform, ROHs shorter than 3 Mb were excluded. (C) Chromosomal distribution of ROH counts. (D) ROH densities at different chromosomal regions. ROH locations are shown as horizontal blue bars on the karyotype plot, where the height of each stacked bar indicates the ROH density. ROH hotspots present in over 10% of patients were observed at 3p21.31p21.1 (20.3%), 11p11.2 (18.2%), 1q21.1q21.3 (17.7%), and 1p33p32.3 (12.0%). Pericentric clusters were found in chromosomes 2, 3, 5, 7, 8, 11, 17, and 20, as indicated by the red triangles.
Next, we analyzed the distributions of ROH counts across chromosomes (Fig. 1C) and their densities at the chromosomal band level (Fig. 1D). The most frequent ROH hotspots, present in over 10% of patients, were located at 3p21.31p21.1 (20.3%), 11p11.2 (18.2%), 1q21.1q21.3 (17.7%), and 1p33p32.3 (12.0%). We also investigated common ROH regions found in at least 1% of the cohort (Table 1). With one exception, the common ROHs had a mean size of <5 Mb.
Table 1. Common regions of homozygosity found in at least 1% of the study cohort.
| Location | Median size (Mb, range) | Count | Proportion (%) |
|---|---|---|---|
| 3p21.31p21.1 | 3.7 (3.0–7.1) | 351 | 20.3 |
| 11p11.2 | 3.7 (3.0–10.8) | 315 | 18.2 |
| 1q21.1q21.3 | 3.2 (3.0–6.6) | 306 | 17.7 |
| 1p33p32.3 | 4.0 (3.0–5.0) | 207 | 12.0 |
| 2q11.1q11.2 | 3.4 (3.0–5.7) | 70 | 4.0 |
| 20q11.21q11.23 | 3.3 (3.0–5.6) | 69 | 4.0 |
| 7q11.21q11.23 | 3.3 (3.0–5.1) | 60 | 3.5 |
| 15q15.1q21.1 | 3.4 (3.0–5.6) | 49 | 2.8 |
| 8q11.1q11.23 | 3.3 (3.0–5.1) | 49 | 2.8 |
| 17q22q23.2 | 3.4 (3.0–4.2) | 48 | 2.8 |
| 12q24.11q24.31 | 3.2 (3.0–4.0) | 34 | 2.0 |
| 17q11.1q11.2 | 3.3 (3.0–4.3) | 27 | 1.6 |
| 17q21.2q21.32 | 3.1 (3.0–5.6) | 27 | 1.6 |
| 10q22.1q22.2 | 3.3 (3.0–3.8) | 26 | 1.5 |
| 2q32.1q32.3 | 3.4 (3.0–4.6) | 26 | 1.5 |
| 6p22.3p22.1 | 3.6 (3.0–4.7) | 26 | 1.5 |
| 15q22.2q22.32 | 3.1 (3.0–3.9) | 25 | 1.4 |
| 6p22.3p21.31 | 5.2 (3.1–11.4) | 24 | 1.4 |
| 16q11.2q12.1 | 3.1 (3.0–4.8) | 21 | 1.2 |
| 3p12.3p12.1 | 3.4 (3.0–4.1) | 21 | 1.2 |
| 5p13.1p12 | 3.2 (3.0–3.9) | 21 | 1.2 |
| 8q23.1q23.3 | 3.4 (3.1–5.2) | 21 | 1.2 |
| 16q22.1q22.3 | 3.4 (3.0–4.6) | 20 | 1.2 |
| 12q21.1q21.33 | 3.3 (3.0–5.1) | 19 | 1.1 |
| 1p36.11p35.2 | 3.2 (3.0–4.0) | 19 | 1.1 |
| 16q21q22.2 | 3.3 (3.0–4.7) | 17 | 1.0 |
| 22q13.1q13.33 | 3.2 (3.0–4.7) | 17 | 1.0 |
| 3q26.1q26.2 | 3.3 (3.0–4.0) | 17 | 1.0 |
| 5q31.2q31.3 | 3.6 (3.0–4.4) | 17 | 1.0 |
Abbreviation: Mb, megabase.
Association between ROHs and recombination
ROHs were frequently detected in the pericentric regions of chromosomes 2, 3, 5, 7, 8, 11, 17, and 20, presumably resulting from low recombination rates in these regions [17]. To further investigate the association between the recombination rate and ROH frequency, we conducted regression analysis of the known recombination rates of each human autosome [15] and the ROH count per chromosome observed in this study (Fig. 2A). The recombination rate correlated negatively with the number of ROHs in each chromosome (R2=0.20, P=0.039). Additionally, regression analysis revealed a weak positive correlation with the ROH frequency (R2=0.025, P=0.011; Fig. 2B). Collectively, these findings suggest that regions with low recombination rates, such as those with high gene densities, are more susceptible to ROH formation.
Fig. 2. Association between regions of homozygosity (ROHs) and recombination rates. (A) The recombination rate of each chromosome showed a significant negative correlation with the number of ROHs identified per chromosome. (B) The ROH frequency was significantly and positively associated with the mean gene density in the corresponding regions.
Abbreviations: cM, centimorgan; Mb, megabase; kb, kilobase.
Classification of ROH patterns and their potential origins
To explore the clinical implications of ROHs, we categorized individuals based on their ROH patterns and potential origins, following ACMG’s technical standards. The individuals were classified into three groups, including those with no detectable ROH (23.3%, N=403), suspected UPD or consanguinity (1.2%, N= 21), or other types of ROHs (75.5%, N=1,307).
We identified two cases of suspected consanguinity, 13 cases of suspected UPD on imprinted chromosomes, and six cases of suspected UPD on non-imprinted chromosomes.
Chromosomal patterns in cases of suspected UPD or consanguinity
To further characterize the genomic signatures associated with suspected UPD or consanguinity, we analyzed the distribution of ROHs in these cases (Fig. 3A). In cases where UPD was suspected, the ROHs spanned entire chromosomes, whole chromosome arms, or specific segmental and pericentric regions within a single chromosome. In some (but not all) cases, these ROHs overlapped with imprinted loci associated with imprinting disorders [18].
Fig. 3. Regions of homozygosity (ROH) distributions in suspected cases of uniparental disomy (UPD) and consanguinity. (A) In suspected cases of UPD, ROHs can span an entire chromosome, a whole chromosome arm, or specific segmental regions (interstitial, pericentric, or terminal). Suspected UPD was identified in all examined imprinted chromosomes (6, 7, 11, 14, 15, and 20). The purple stars represent known imprinted loci associated with imprinting disorders. The case numbers from Table 2 are indicated next to the corresponding ROH range markers. (B) In the two suspected cases of consanguinity (Case 20 and Case 21 from Table 2), multiple ROHs were detected across various chromosomes.
Abbreviations: s-UPD on non-imp., suspected uniparental disomy on non-imprinted chromosomes; s-UPD on imp., suspected uniparental disomy on imprinted chromosomes; s-Consang., suspected consanguinity.
All imprinted chromosomes (6, 7, 11, 14, 15, and 20) were affected; this included four cases of interstitial segmental ROHs (chromosomes 6, 15, and 20), four cases of pericentric segmental ROHs (chromosome 11), two cases of whole-chromosome UPD (chromosomes 6 and 7), one case of a whole-arm-level ROH (chromosome 20), and one case of a terminal segmental ROH (chromosome 6). In contrast, the two cases of suspected consanguinity exhibited multiple ROHs distributed across multiple chromosomes (Fig. 3B).
Association between ROHs and CNVs
CNV formation is influenced by recombination events; therefore, we investigated the potential relationships between ROHs and CNVs. Our analysis revealed no significant difference in CNV counts across different ROH categories (Fig. 4A).
Fig. 4. Pathogenicity of concurrent copy number variations (CNVs) with different types of regions of homozygosity (ROHs). (A) No significant difference was observed in the number of concurrent CNVs among different ROH types. (B) Cases of suspected consanguinity and suspected uniparental disomy (UPD) on imprinted chromosomes did not exhibit any CNVs classified as pathogenic, likely pathogenic, or of uncertain significance, unlike other types of ROHs.
Abbreviations: ns, not significant; s-Consang., suspected consanguinity; s-UPD on imp., suspected uniparental disomy on imprinted chromosomes; s-UPD on non-imp, suspected uniparental disomy on non-imprinted chromosomes.
However, when evaluating the pathogenicity of concurrent CNVs, we found that cases of suspected consanguinity or UPD on imprinted chromosomes lacked clinically significant CNVs (pathogenic or likely pathogenic variants). In contrast, clinically significant CNVs were identified in 11.9% of individuals with no ROHs, 16.7% with ROHs on non-imprinted chromosomes, and 22.6% with other ROHs (Fig. 4B).
Clinical diagnosis of suspected cases of consanguinity and UPD
Finally, we reviewed the clinical information from 13 patients with suspected UPD or consanguinity (Table 2). Among the 13 cases of suspected UPD on imprinted chromosomes, eight patients were diagnosed with the corresponding imprinting disorders through clinical evaluation by medical geneticists and confirmatory diagnostic testing, where available. Specifically, Case 4 was diagnosed as having transient neonatal diabetes mellitus type 1 based on phenotypic features, including symmetrical intrauterine growth restriction. Case 5 was diagnosed as having Silver–Russell syndrome-2 based on phenotypic features, as well as methylation-specific PCR and SGCE restriction fragment length polymorphism analyses, with the results indicating maternal UPD of chromosome 7. Case 7 was diagnosed as having Beckwith–Wiedemann syndrome based on phenotypic features, including facial dysmorphism and omphalocele. Case 9 was diagnosed as having Silver–Russell syndrome based on phenotypic features, including small for gestational age and short stature. Case 10 was diagnosed as having Kagami–Ogata syndrome based on phenotypic features, including omphalocele, facial dysmorphism, and a karyotype showing a Robertsonian translocation involving chromosomes 13 and 14 inherited from the father. Case 11 was diagnosed as having Prader–Willi syndrome based on phenotypic features and methylation-specific PCR analysis of maternal SNRPN-methylation patterns. Case 12 was diagnosed as having Mulchandani–Bhoj–Conlin syndrome based on phenotypic features and parental CMA testing indicative of maternal UPD of chromosome 20. Lastly, Case 13 was diagnosed as having Mulchandani–Bhoj–Conlin syndrome based on phenotypic features, including short stature and failure to thrive.
Table 2. Suspected cases of UPD or consanguinity.
| Case No. | Age | Sex | ROH type | Chromosome(s) | Location(s) | Total size (Mb) | Diagnosis/phenotype* | Concurrent CNV† | Ethnicity |
|---|---|---|---|---|---|---|---|---|---|
| 1 | 5 yrs | F | s-UPD on imp. | 6 | Interstitial (p22.3p21.31) | 11.4 | Autism, mental retardation, dysmorphism | Benign | Korean |
| 2 | 63 days | F | s-UPD on imp. | 6 | Interstitial (q13q14.1) | 11.8 | Severe tricuspid valve regurgitation, tricuspid dysplasia | Benign | Korean |
| 3 | 9 yrs | F | s-UPD on imp. | 6 | Terminal (p25.3p21.31) | 35.4 | Systemic sclerosis, severe aplastic anemia | Benign | Korean |
| 4 | 9 days | F | s-UPD on imp. | 6 | Whole chromosome | 166.7 | Diabetes mellitus, transient neonatal 1‡ | Benign | Korean |
| 5 | 2 months | F | s-UPD on imp. | 7 | Whole chromosome | 154.6 | Silver–Russell syndrome‡ | Benign | Korean |
| 6 | 8 yrs | M | s-UPD on imp. | 11 | Pericentric (p11.2q11.12) | 10.8 | NEDBEH: heterozygous pathogenic variants of RERE1 on 1p36.23 | Benign | Korean |
| 7 | 4 days | M | s-UPD on imp. | 11 | Pericentric (p11.2q12.1) | 11.1 | Beckwith–Wiedemann syndrome‡ | Benign | Korean |
| 8 | 36 yrs | M | s-UPD on imp. | 11 | Pericentric (p11.2q12.1) | 12.3 | Absence of finger and toe, Brachydactyly | Benign | Korean |
| 9 | 5 yrs | M | s-UPD on imp. | 11 | Pericentric (p11.2q12.1) | 12.6 | Silver–Russell syndrome‡ combined with hypophosphatasia, childhood (heterozygous pathogenic variants of ALPL on 1p36.1) | Benign | Korean |
| 10 | 13 days | M | s-UPD on imp. | 14 | Terminal (q31.3q32.33) | 18.6 | Kagami–Ogata syndrome‡ | Benign | Korean |
| 11 | 2 months | F | s-UPD on imp. | 15 | Interstitial (q12q22.2) | 31.8 | Prader–Willi syndrome‡ | Benign | Korean |
| 12 | 3 yrs | M | s-UPD on imp. | 20 | Whole q-arm | 33.5 | Mulchandani–Bhoj–Conlin syndrome‡ | Benign | Korean |
| 13 | 4 yrs | F | s-UPD on imp. | 20 | Interstitial (p11.23q13.31) | 35.5 | Mulchandani–Bhoj–Conlin syndrome‡ | Benign | Korean |
| 14 | 3 yrs | M | s-UPD on non-imp. | 2 | Near-whole chromosome | 224.4 | Marker chromosome of 2p12q11.2 (19.0 Mb) gain | Pathogenic (2p12q11.2 gain, 19.0 Mb) | Korean |
| 15 | 19 yrs | M | s-UPD on non-imp. | 2 | Whole chromosome | 242.8 | Skeletal dysplasia, restrictive lung disease | Benign | Korean |
| 16 | 2 months | F | s-UPD on non-imp. | 3 | Interstitial (p24.3p22.2, q13.31q21.1) | 22.4 | Congenital hypothyroidism, ventricular septal defect, ventriculomegaly | Benign | Filipino |
| 17 | 1 day | M | s-UPD on non-imp. | 5 | Interstitial (p15.32q11.2, q34q35.3) | 62.8 | Trisomy 5 on prenatal CVS karyotyping, single umbilical artery, duplex kidney, umbilical cord varix | VUS (2p25.3 gain, 335 kb; 13q33.1 gain, 948 kb) | Korean |
| 18 | 1 yr, 8 months | F | s-UPD on non-imp. | 8 | Whole p-arm | 42.2 | Developmental delay, prematurity, and spastic lower legs | Benign | Korean |
| 19 | 5 yrs | M | s-UPD on non-imp. | 17 | Whole q-arm | 55.4 | Homozygous pathogenic variants of NF1 (on 17q11.2) | Benign | Korean |
| 20 | 3 yrs | F | s-Consang. | 1, 3, 5, 8, 14, 18 | 139.4 | ROH percentage: 4.8%: above 4th degree | Benign | Indian | |
| 21 | 10 yrs | M | s-Consang. | 1, 2, 3, 6, 7, 9, 10, 11, 12, 13, 17, 18 | 421.4 | ROH percentage: 14.6%: above 2nd degree | Benign | Middle Eastern |
*The genetic cause or corresponding diagnosis is provided; if unknown, the phenotype is described.
†For concurrent CNVs interpreted as pathogenic, likely pathogenic, or VUS, the location and size are indicated in parentheses.
‡Imprinting disorders associated with the affected chromosomes are indicated.
Abbreviations: UPD, uniparental disomy; ROH: region of homozygosity; Mb, megabase; CNV: copy number variation; F, female; s-UPD on imp., suspected uniparental disomy on imprinted chromosomes; s-UPD on non-imp., suspected uniparental disomy on non-imprinted chromosomes; M, male; NEDBEH, Neurodevelopmental disorder with or without anomalies of the brain, eye, or heart; CVS, chorionic villus sampling; VUS, variant of uncertain significance; s-Consang., suspected consanguinity.
In one case of suspected UPD on a non-imprinted chromosome, whole-genome sequencing revealed homozygous pathogenic NF1 variants within the ROH. Both patients with suspected consanguinity had ancestries traced to the Middle East or South Asia, and the degree of consanguinity could be estimated based on the proportion of the genome covered by the ROHs.
DISCUSSION
Interpreting ROHs in clinical laboratories has been challenging, given the limited availability of population-specific data and well-established criteria. Our findings help address these challenges by providing real-world reference data and demonstrating the clinical utility of standardized criteria for detecting UPD and consanguinity, reinforcing the importance of systematic ROH evaluations.
Notably, we identified frequently shared ROH regions, with the most highly conserved regions located at 3p21.31p21.1 (20.3%), 11p11.2 (18.2%), 1q21.1q21.3 (17.7%), and 1p33p32.3 (12.0%). These results align with previous findings for other ethnic groups using a 3 Mb ROH threshold [11, 12], which identified similar hotspots, albeit at varying frequencies, supporting the functional significance of these regions. For example, studying a large Middle Eastern cohort of 13,383 predominantly prenatal CMA samples revealed ROHs in over 5% of participants at 15q15.1q21.1 (16.3%), 3p21.31p21.2 (16.6%), 3p21.31p21.1 (14.1%), 11p11.2p11.12 (10.7%), and 1p33p32.3 (6.3%) [11]. Similarly, the results of a Brazilian study of 430 neurodevelopmental disorder patients identified common ROHs at 16p11.2p11.1 (49%), 1q21.2q21.3 (21%), 11p11.2p11.12 (19%), and 3p21.31p21.2 (16%) [12]. Although variations across studies can arise due to differences in diagnostic platforms [8], the recurrence of these ROH regions highlights evolutionary forces (such as selective sweeps, population bottlenecks, and reduced recombination) that collectively drive their persistence in these genomic segments [9, 13, 14]. Concordantly, we observed a negative correlation between the ROH frequency and recombination rate at the chromosomal level and a positive correlation with the gene density, further emphasizing the non-random nature of ROH distributions.
Nonetheless, establishing uniform ROH size criteria for detecting potentially pathogenic cases remains challenging, given the genetic heterogeneity across various ethnic groups. For instance, the results of a study conducted in Singapore showed that individuals in an Indian population exhibited fewer and shorter ROHs than those in Chinese and Malay populations [13]. Similarly, total ROH percentages and sizes varied significantly among different ethnic groups in the Middle East [19]. In ethnically diverse populations, tailored approaches–rather than fixed absolute size thresholds–are warranted to ensure both the accuracy and clinical relevance of ROH reporting [19].
However, in populations such as East Asian and European populations–where only baseline levels of shared ancestry are typically observed [20, 21]–the ROH patterns exhibit less variation between individuals. Consequently, applying universal ROH size criteria for detecting pathogenic variants is more feasible. In this study, using ACMG’s technical standards, we identified a 1.1% frequency of suspected UPD. Previously, UPD was reported to occur in approximately one in 2,000 live births (0.05%) in the general population [22]. However, the incidence was reported to be as high as one in 176 live births (0.57%) among individuals with developmental delays [23]. Our 1.1% detection rate is particularly representative of clinical settings, as it was determined solely based on the ROH size–without parental testing–in patients undergoing CMA as a first-tier test. This frequency also aligns with findings from a large prenatal CMA study, for which a 1.8% rate of suspected UPD was reported based on a 10 Mb threshold [24].
Additionally, we identified two suspected cases of consanguinity in individuals from ethnic backgrounds where consanguineous marriages are prevalent, emphasizing the role of demographic and cultural factors in ROH interpretations–even in countries like Korea, where consanguinity is legally restricted.
Our findings further indicate that large ROHs on imprinted chromosomes can directly point to UPD as an underlying cause of disease. Specifically, among individuals with large ROHs on imprinted chromosomes–despite the absence of pathogenic CNVs–eight of 13 were diagnosed as having imprinting disorders upon additional testing and phenotype assessment. In contrast, we found that 11.9%, 16.7%, and 22.6% of individuals without ROHs, with ROHs on non-imprinted chromosomes, or other ROHs, respectively, had pathogenic or likely pathogenic CNVs. Even with some cases of suspected UPD on imprinted chromosomes that did not include known imprinted regions [18], imprinting disorders corresponding to the affected chromosomes were eventually diagnosed (e.g., Case 7 and Case 9 in Table 2). These observations suggest that mixed isodisomy/heterodisomy UPD, arising from meiotic recombination, may also contribute to disease pathogenesis. Overall, our results underscore the need for thoroughly evaluating cases of ROHs on imprinted chromosomes, as such assessments can uncover epigenetic disease mechanisms undetectable by standard CNV analyses.
A limitation of this study is the 3 Mb resolution threshold used for ROH detection, which is suitable for diagnostic purposes but may overlook smaller ROHs with potential clinical relevance. However, given that ROHs up to 4 Mb are commonly observed in outbred individuals of European descent [3], the likelihood of smaller ROHs being pathogenic remains low. We found that almost all common ROHs were shorter than 5 Mb. Although most clinical laboratories define ≥3–5 Mb as the threshold for classifying ROHs as clinically significant, setting the reporting criteria at ≥5 Mb would be more practical. Nonetheless, the increasing adoption of whole-genome sequencing in clinical practice is rendering testing for pathogenic homozygous variants within smaller ROHs more feasible. Such analyses may offer deeper insights into disease mechanisms and further clarify the role of ROHs in genetic disorders.
In conclusion, our findings provide real-world data for common ROH patterns in a Korean population, serving as a valuable reference for improving clinical interpretation by minimizing the risk of misclassifying benign variants. Our findings highlight the importance of standardized criteria, such as ACMG’s technical standards, in reliably identifying suspected UPD and consanguinity, which reinforces the need for integrating these standards into routine postnatal CMA interpretations. Future studies with larger population-specific ROH datasets are warranted to further enhance the accuracy, consistency, and reliability of ROH interpretations in various clinical contexts.
ACKNOWLEDGEMENTS
None.
Footnotes
AUTHOR CONTRIBUTIONS
Kim J contributed to the conception, methodology, investigation, visualization, and writing the original draft; Seo EJ contributed to the conception, methodology, investigation, funding acquisition, project administration, supervision, and writing (review and editing); and Min S and Seol CA contributed to the investigation. All authors read and approved the final manuscript.
CONFLICTS OF INTEREST
None declared.
RESEARCH FUNDING
None declared.
REFERENCES
- 1.South ST, Lee C, Lamb AN, Higgins AW, Kearney HM Working Group for the American College of Medical Genetics and Genomics Laboratory Quality Assurance Committee, author. ACMG standards and guidelines for constitutional cytogenomic microarray analysis, including postnatal and prenatal applications: revision 2013. Genet Med. 2013;15:901–9. doi: 10.1038/gim.2013.129. [DOI] [PubMed] [Google Scholar]
- 2.Wang JC, Ross L, Mahon LW, Owen R, Hemmat M, Wang BT, et al. Regions of homozygosity identified by oligonucleotide SNP arrays: evaluating the incidence and clinical utility. Eur J Hum Genet. 2015;23:663–71. doi: 10.1038/ejhg.2014.153. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3.McQuillan R, Leutenegger AL, Abdel-Rahman R, Franklin CS, Pericic M, Barac-Lauc L, et al. Runs of homozygosity in European populations. Am J Hum Genet. 2008;83:359–72. doi: 10.1016/j.ajhg.2008.08.007. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4.Sund KL, Zimmerman SL, Thomas C, Mitchell AL, Prada CE, Grote L, et al. Regions of homozygosity identified by SNP microarray analysis aid in the diagnosis of autosomal recessive disease and incidentally detect parental blood relationships. Genet Med. 2013;15:70–8. doi: 10.1038/gim.2012.94. [DOI] [PubMed] [Google Scholar]
- 5.Hoppman N, Rumilla K, Lauer E, Kearney H, Thorland E. Patterns of homozygosity in patients with uniparental disomy: detection rate and suggested reporting thresholds for SNP microarrays. Genet Med. 2018;20:1522–7. doi: 10.1038/gim.2018.24. [DOI] [PubMed] [Google Scholar]
- 6.Papenhausen P, Schwartz S, Risheg H, Keitges E, Gadi I, Burnside RD, et al. UPD detection using homozygosity profiling with a SNP genotyping microarray. Am J Med Genet A. 2011;155:757–68. doi: 10.1002/ajmg.a.33939. [DOI] [PubMed] [Google Scholar]
- 7.Del Gaudio D, Shinawi M, Astbury C, Tayeh MK, Deak KL, Raca G, et al. Diagnostic testing for uniparental disomy: a point to consider statement from the American College of Medical Genetics and Genomics (ACMG) Genet Med. 2020;22:1133–41. doi: 10.1038/s41436-020-0782-9. [DOI] [PubMed] [Google Scholar]
- 8.Wang JC, Radcliff J, Coe SJ, Mahon LW. Effects of platforms, size filter cutoffs, and targeted regions of cytogenomic microarray on detection of copy number variants and uniparental disomy in prenatal diagnosis: results from 5026 pregnancies. Prenat Diagn. 2019;39:137–56. doi: 10.1002/pd.5375. [DOI] [PubMed] [Google Scholar]
- 9.Matalonga L, Laurie S, Papakonstantinou A, Piscia D, Mereu E, Bullich G, et al. Improved diagnosis of rare disease patients through systematic detection of runs of homozygosity. J Mol Diagn. 2020;22:1205–15. doi: 10.1016/j.jmoldx.2020.06.008. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10.Gonzales PR, Andersen EF, Brown TR, Horner VL, Horwitz J, Rehder CW, et al. Interpretation and reporting of large regions of homozygosity and suspected consanguinity/uniparental disomy, 2021 revision: a technical standard of the American College of Medical Genetics and Genomics (ACMG) Genet Med. 2022;24:255–61. doi: 10.1016/j.gim.2021.10.004. [DOI] [PubMed] [Google Scholar]
- 11.Sagi-Dain L, Levy M, Matar R, Kahana S, Agmon-Fishman I, Klein C, et al. Exploring the human genomic landscape: patterns of common homozygosity regions in a large middle eastern cohort. Hum Mol Genet. 2024;33:1908–15. doi: 10.1093/hmg/ddae123. [DOI] [PubMed] [Google Scholar]
- 12.Chaves TF, Oliveira LF, Ocampos M, Barbato IT, de Luca GR, Barbato Filho JH, et al. Long contiguous stretches of homozygosity detected by chromosomal microarrays (CMA) in patients with neurodevelopmental disorders in the South of Brazil. BMC Med Genomics. 2019;12:50. doi: 10.1186/s12920-019-0496-5.f938e8a1253c47ea91e1b74323af03be [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13.Teo SM, Ku CS, Salim A, Naidoo N, Chia KS, Pawitan Y. Regions of homozygosity in three southeast Asian populations. J Hum Genet. 2012;57:101–8. doi: 10.1038/jhg.2011.132. [DOI] [PubMed] [Google Scholar]
- 14.Pemberton TJ, Absher D, Feldman MW, Myers RM, Rosenberg NA, Li JZ. Genomic patterns of homozygosity in worldwide human populations. Am J Hum Genet. 2012;91:275–92. doi: 10.1016/j.ajhg.2012.06.014. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15.Jensen-Seaman MI, Furey TS, Payseur BA, Lu Y, Roskin KM, Chen CF, et al. Comparative recombination rates in the rat, mouse, and human genomes. Genome Res. 2004;14:528–38. doi: 10.1101/gr.1970304. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16.Riggs ER, Andersen EF, Cherry AM, Kantarci S, Kearney H, Patel A, et al. Technical standards for the interpretation and reporting of constitutional copy-number variants: a joint consensus recommendation of the American College of Medical Genetics and Genomics (ACMG) and the Clinical Genome Resource (ClinGen) Genet Med. 2020;22:245–57. doi: 10.1038/s41436-019-0686-8. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17.Nambiar M, Smith GR. Repression of harmful meiotic recombination in centromeric regions. Semin Cell Dev Biol. 2016;54:188–97. doi: 10.1016/j.semcdb.2016.01.042. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18.Kearney HM, Kearney JB, Conlin LK. Diagnostic implications of excessive homozygosity detected by SNP-based microarrays: consanguinity, uniparental disomy, and recessive single-gene mutations. Clin Lab Med. 2011;31:595–613. doi: 10.1016/j.cll.2011.08.003. [DOI] [PubMed] [Google Scholar]
- 19.Maya I, Levy M, Matar R, Kahana S, Agmon-Fishman I, Klein C, et al. Regions of homozygocity size patterns among diverse ethnic groups in Israel: toward tailored diagnostic reporting thresholds. Am J Med Genet A. 2024;194:e63839. doi: 10.1002/ajmg.a.63839. [DOI] [PubMed] [Google Scholar]
- 20.Kirin M, McQuillan R, Franklin CS, Campbell H, McKeigue PM, Wilson JF. Genomic runs of homozygosity record population history and consanguinity. PLoS One. 2010;5:e13996. doi: 10.1371/journal.pone.0013996.97fc13310262473f87ef596312f05c6e [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21.Ceballos FC, Joshi PK, Clark DW, Ramsay M, Wilson JF. Runs of homozygosity: windows into population history and trait architecture. Nat Rev Genet. 2018;19:220–34. doi: 10.1038/nrg.2017.109. [DOI] [PubMed] [Google Scholar]
- 22.Nakka P, Pattillo Smith S, O'Donnell-Luria AH, McManus KF, Mountain JL, et al. 23andMe Research Team, author. Characterization of prevalence and health consequences of uniparental disomy in four million individuals from the general population. Am J Hum Genet. 2019;105:921–32. doi: 10.1016/j.ajhg.2019.09.016. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23.King DA, Fitzgerald TW, Miller R, Canham N, Clayton-Smith J, Johnson D, et al. A novel method for detecting uniparental disomy from trio genotypes identifies a significant excess in children with developmental disorders. Genome Res. 2014;24:673–87. doi: 10.1101/gr.160465.113. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24.Xu C, Li M, Gu T, Xie F, Zhang Y, Wang D, et al. Chromosomal microarray analysis for prenatal diagnosis of uniparental disomy: a retrospective study. Mol Cytogenet. 2024;17:3. doi: 10.1186/s13039-023-00668-8.107cfda82f7d4ccf93b7498e80d5d9b3 [DOI] [PMC free article] [PubMed] [Google Scholar]