Implications of uniparental disomy in forensic kinship testing: A case study of paternal isodisomy on chromosome 3

TL;DR

This case study shows how uniparental disomy can cause misleading results in forensic DNA testing and suggests ways to detect and handle such cases.

Contribution

The study demonstrates the forensic impact of paternal isodisomy on chromosome 3 and proposes strategies to detect and interpret such cases.

Findings

Complete paternal isodisomy on chromosome 3 was confirmed using STR sequencing and SNP microarray testing.

UPD caused inconclusive maternal and secondary relationship analyses despite conclusive paternity results.

Software alerts are recommended to flag potential UPD patterns in forensic testing.

Abstract

In typical inheritance, a child receives one chromosome of each pair from each parent. In rare cases, however, both chromosomes may be inherited from the same parent, a phenomenon known as uniparental disomy (UPD). In forensic kinship testing, UPD can lead to Mendelian inconsistencies between parent and child, increasing the risk of inconclusive or erroneous interpretations. In this case study, inconsistencies between the mother and child during paternity testing prompted further investigation. Parentage was confirmed (probability of maternity and paternity >99.99%), using autosomal short tandem repeat (STR) typing. Additional analyses were performed, including STR sequencing via next‐generation sequencing (NGS) and single nucleotide polymorphism (SNP) microarray testing across all trio samples. A total of 4 STRs and 273 SNPs on Chromosome 3 were examined, confirming complete paternal isodisomy in the child. Alternative kinship scenarios were then evaluated to assess the impact of UPD on relationship testing. While paternity results remained conclusive, maternal and secondary relationship analyses produced inconclusive and false outcomes, even when up to 42 STR loci were included. This study highlights the importance of recognizing UPD and its genotypic features in forensic casework. To mitigate the risk of misinterpretation, forensic scientists should remain vigilant for multiple inconsistencies restricted to a single chromosome, supported by the implementation of software alerts designed to flag such patterns.

Figures

Electropherograms showing allele profiles for D3S1358 (A), D3S1744 (B), D3S3045 (C), and D10S1435 (D). For each locus, profiles are shown for the mother (top), child (middle), and alleged father (bottom).

Tables

TABLE 1. Sequence motifs and alleles for STR loci on chromosome 3 determined using the GlobalFiler NGS STR Panel.

Locus	Mother		Child		Alleged father		Support for UPD
	Sequence	Allele	Sequence	Allele	Sequence	Allele
D3S1358	[TCTA]1 [TCTG]1 [TCTA]13	15	[TCTA]1 [TCTG]2 [TCTA]11	14	[TCTA]1 [TCTG]2 [TCTA]11	14	Requires multiple maternal mutations (variation across two separate repeat units) to match child genotype
	[TCTA]1 [TCTG]3 [TCTA]13	17
D3S4529	[ATCT]7 ATTT[ATCT]4	12	[ATCT]10 ATTT[ATCT]4	15	[ATCT]8 ATTT[ATCT]4	13	Requires at least a two‐step maternal mutation
	[ATCT]8 ATTT[ATCT]4	13			[ATCT]10 ATTT[ATCT]4	15

TABLE 2. SNPs on chromosome 3 showing mismatching genotypes between mother and child.

SNP	Mother	Child	Alleged father	Chromosomal location
rs2030874	TT	CC	CT	5506255
rs2254298	GG	AA	AG	8760542
rs12632942	AA	GG	GG	38723507
rs10212536	AA	GG	AG	41785534
rs9815354	GG	AA	AG	41871159
rs1800023	GG	AA	AA	46370817
rs333	II	DD	DI	46373456
rs4607103	TT	CC	CC	64726228
rs2371767	CC	GG	CG	64732582
rs984038	CC	TT	TT	65776577
rs6548616	TT	CC	CT	79350425
rs12495178	TT	CC	CT	85836927
rs1465648	CC	TT	CT	87901079
rs12639347	TT	CC	CC	105154433
rs7630522	CC	TT	TT	107434241
rs2134655	CC	TT	TT	114139354
rs17251221	AA	GG	AG	122274400
rs1801725	GG	TT	GT	122284910
rs17265703	AA	GG	AG	122329797
rs820371	CC	TT	CT	123685864
rs28497577	GG	TT	GT	123793780
rs2276731	TT	CC	CT	126155545
rs2977562	GG	AA	AG	128387424
rs1511412	GG	AA	AG	138994862
rs940187	CC	TT	CT	139122751
rs868767	AA	GG	GG	141658488
rs3772622	TT	CC	CT	148717966
rs344081	TT	CC	CT	156838195
rs568408	GG	AA	AG	159995680
rs448378	AA	GG	AG	169383111
rs1344555	TT	CC	CC	169582431
rs2922126	AA	TT	TT	172449471
rs4402960	GG	TT	GT	185793899
rs1470579	AA	CC	AC	185811292
rs720390	GG	AA	AG	185830895
rs6763645	AA	GG	AG	185990053
rs4917	TT	CC	CT	186619924
rs1042464	AA	TT	TT	186677783
rs266729	GG	CC	CC	186841685
rs182052	AA	GG	GG	186842993
rs1501299	GG	TT	GT	186853334
rs1464510	CC	AA	AA	188394766
rs779306	TT	CC	CC	191774566

TABLE 3. Combined likelihood ratios (CLR) from kinship testing using different STR panels.

Test type	15 STRs (Identifiler)	23 STRs (VeriFiler)	42 STRs (VeriFiler + 27comp)
Trio paternity	1.96 × 107	1.36 × 1012	1.18 × 1021
Duo paternity	4.97 × 104	6.19 × 108	2.28 × 1015
Trio maternity	1.28 × 105	1.36 × 1012	1.10 × 106
Duo maternity	325.20	2.62E × 105	2.12
Duo father vs. uncle	64.71	1249.92	2.82 × 105
Trio father vs. uncle	481.31	1.36 × 104	1.91 × 107
Duo mother vs. aunt	0.16	1.27	0.00000019
Trio mother vs. aunt	1.02	14.00	0.000028

Full text

Highlights

• Confirms paternal isodisomy of chromosome 3 using short tandem repeat (STR), next‐generation sequencing (NGS‐STR), and single nucleotide polymorphism (SNP) forensic DNA analysis.
• Demonstrates UPD can cause false exclusions in mother–child relationship testing.
• Shows increased STR loci may worsen interpretation in secondary kinship testing with uniparental disomy (UPD).
• Recommends software alerts for inconsistencies confined to a single chromosome.
• Emphasizes need for forensic awareness of UPD to prevent misinterpretation in casework.

INTRODUCTION

1

In typical human inheritance, a child receives one copy of each chromosome from the mother and one from the father, resulting in two copies of each of the 22 autosomal chromosomes. In rare cases, however, both copies of a chromosome may be inherited from the same parent, deviating from expected Mendelian inheritance patterns. This phenomenon is referred to as uniparental disomy (UPD). More specifically, isodisomy describes the inheritance of two identical homologous chromosomes from one parent, whereas heterodisomy refers to the inheritance of two different homologues from the same parent. These outcomes may arise from meiosis errors or postzygotic chromosomal rescue mechanisms [1, 2].

While UPD can cause a range of clinical conditions, including autosomal recessive disorders and imprinting syndromes, its implications extend beyond medicine. In forensic science, the presence of UPD can influence the outcomes of kinship testing. For chromosomes affected by UPD, the assumption that one allele at each locus is inherited from each parent no longer holds true, complicating interpretation. This challenge becomes more pronounced as the number of loci per chromosome increases, raising the likelihood of incompatibilities and the risk of misinterpretation, including false exclusions [3].

Short tandem repeat (STR) typing, using fragment analysis via capillary electrophoresis, is currently the preferred method for kinship testing because of its cost‐effectiveness, informativeness, and established reliability [4]. However, this approach may be limited in cases involving mismatches at loci affected by UPD. Next‐generation sequencing (NGS) can provide additional insight by determining sequence‐level detail at STR loci of interest [5, 6]. DNA microarrays can also be used to analyze numerous single nucleotide polymorphisms (SNPs), which are particularly useful for identifying isodisomy and heterodisomy due to their abundance and even genomic distribution [7, 8]. While detecting heterodisomy with SNPs alone can be unreliable owing to low variability, combining STR and SNP data is recommended to improve interpretation [9, 10].

This case study aims to explore the complications that UPD introduces into kinship testing. It examines scenarios that may lead to inconclusive findings or misinterpretation and proposes strategies to mitigate these challenges. By raising awareness of UPD in the forensic context, the study also seeks to improve detection and strengthen the reliability of relationship testing outcomes.

MATERIALS AND METHODS

2

Study samples

2.1

Three samples from individuals of British ancestry forming a family trio (mother, female child, and alleged father) were collected via buccal swabs for the purpose of paternity testing. Parentage was assessed using the VeriFiler Express PCR kit (Thermo Fisher Scientific, Massachusetts, United States). A single‐step mismatch at locus D3S1358 was treated as a mutation, in accordance with recommendations from the forensic science regulator (FSR) [11]. The resulting likelihood ratios exceeded the commonly accepted threshold [12]. During a study investigating the challenges of single‐parent kinship testing [13], the presence of UPD was suspected, and these samples were subsequently re‐analyzed for the present report.

STR typing

2.2

DNA extraction was carried out using Prep‐n‐Go Buffer (Thermo Fisher Scientific) [14].

Quantification was performed with the Quantifiler Trio DNA Quantification Kit (Thermo Fisher Scientific) and the QuantStudio 5 Real‐Time PCR System (384‐well format) (Thermo Fisher Scientific), following the respective user manuals. The DNA concentrations obtained were 0.04 ng/μL for the mother, 0.06 ng/μL for the child, and 0.30 ng/μL for the alleged father.

All PCRs were performed using direct amplification without normalization. STR typing at 23 loci was carried out using the VeriFiler Express PCR kit (Thermo Fisher Scientific). Additional amplification was undertaken with the SureID 27comp PCR kit (Ningbo Health Gene Technologies, Ningbo, China), which added 19 further loci for improved discrimination (D18S1364, D13S325, D3S3045, D5S2800, D9S1122, D6S477, D3S1744, D11S2368, D21S2055, D10S1435, D20S482, D8S1132, D7S3048, D19S253, D17S1301, D22GATA198B05, D6S474, D14S1434, D15S659, D4S236) and six overlapping loci with the VeriFiler Express kit (D1S1656, D22S1045, D12S391, D2S441, D10S1248, and D16S539) to verify concordance. Both kits were applied using a reduced‐volume protocol [15] on the GeneAmp PCR System 9700 thermocycler (Thermo Fisher Scientific). The DNA controls supplied with each kit were used for quality assurance.

Capillary electrophoresis was performed on the 3500xL Genetic Analyzer (Thermo Fisher Scientific). STR electropherograms were analyzed using GeneMapper ID‐X v1.6 software (Thermo Fisher Scientific) to identify Mendelian inconsistencies. Analysis thresholds were determined through prior internal validation [15].

NGS

‐ STR analysis

2.3

To confirm identical allele sequences at STR loci between the child and the alleged father, NGS was performed on 31 autosomal STRs using the Precision ID GlobalFiler NGS STR Panel v2 (Thermo Fisher Scientific). Library preparation was carried out using the HID Ion Chef System (Thermo Fisher Scientific), and sequencing was performed on the HID Ion GeneStudio S5 (Thermo Fisher Scientific). Data were analyzed using Converge software with the NGS Application v1.3 (Thermo Fisher Scientific).

SNP microarray testing

2.4

SNP microarray analysis was conducted to confirm homozygosity of chromosome 3 in the child and to identify Mendelian inconsistencies between the child and both parents. Samples were processed using the Rita Infinium iSelect Custom Genotyping Array (Illumina, California, United States) [16], following the Infinium HTS Assay Reference Guide. A total of 8110 SNPs were analyzed before quality filtering, including 277 located on chromosome 3. Automated processing was undertaken using the Freedom EVO platform (Tecan, Männedorf, Switzerland), and raw data were acquired with the iScan System (Illumina). Analysis was performed using GenomeStudio 2.0 software (Illumina).

Evaluation of alternative relationship scenarios

2.5

To evaluate the potential disruption caused by UPD in relationship testing, alternative kinship scenarios were assessed. As both paternity and maternity had been confirmed, additional hypotheses such as parentage versus aunt/uncle relationships were tested. Single‐parent versions of each test were also included to assess the heightened impact of UPD in duo analyses.

Combined likelihood ratios (CLRs) were calculated using Converge software, which allows the calculation of kinship scenarios through the manual drawing of pedigrees. Internal validation was carried out prior to this study, comparing manually calculated likelihood ratios with those generated by Converge across a range of kinship scenarios (from first‐ to third‐degree relatives). All STR loci from both kits were included in the calculation, except for the syntenic loci D21S2055 and D18S1364 from the SureID 27comp kit, which were excluded to reduce linkage bias with Penta D and D18S51, respectively, in the VeriFiler Express kit [17].

Allele frequencies for the reference population were obtained from Perry et al. [18]. The two‐phase mutation model, recommended for STR loci because it accounts for unlimited allele possibilities, was applied [19]. Locus‐specific mutation rates were taken from the 2021 American Association of Blood Banks report [20]. Where unavailable, default rates of 0.001 (paternal) and 0.0003 (maternal) were used [21]. In Converge, Identifiler STR markers (15 autosomal loci) were established using calls from the VeriFiler Express kit.

Acceptance thresholds were set at 10,000 CLR for parentage and 100 CLR for secondary relationships, based on in‐house validations and literature [12].

RESULTS

3

STR data analysis

3.1

Paternity testing on the three samples using the VeriFiler Express kit revealed no inconsistencies between the alleles of the child and the alleged father, with all loci conforming to expected Mendelian inheritance patterns. However, a potential one‐step deletion was observed between the mother and child at locus D3S1358, as shown in Figure 1A.

Electropherograms showing allele profiles for D3S1358 (A), D3S1744 (B), D3S3045 (C), and D10S1435 (D). For each locus, profiles are shown for the mother (top), child (middle), and alleged father (bottom).

To further investigate, additional loci were examined using the SureID 27comp kit. This is a common practice in many laboratories, where additional loci are frequently analyzed to enhance discrimination power, particularly in complex or inconclusive relationship cases [22]. Although a core set of 20 STR loci, in line with the FBI's combined DNA index system (CODIS), is included in many commercial forensic multiplex STR kits [23], including the VeriFiler Express kit, laboratories often extend beyond this core set.

This analysis revealed two further inconsistencies between the mother and child: a potential one‐step deletion at D3S1744 (Figure 1B) and a two‐step insertion at D3S3045 (Figure 1C).

Expanding the analysis, the child was homozygous at all STR loci located on chromosome 3, with no inconsistencies observed between the mother and child at loci on other chromosomes. In contrast, heterozygosity was identified at multiple loci elsewhere, such as D10S1435 on chromosome 10 (Figure 1D).

NGS‐STR data analysis

3.2

NGS analysis confirmed the allelic calls for 26 autosomal STRs previously profiled using the VeriFiler Express and SureID 27comp kits, including D3S1358, demonstrating full concordance. Analysis of the five additional loci included in the GlobalFiler NGS STR Panel identified one more mismatch at D3S4529. All other loci located outside chromosome 3 exhibited Mendelian consistencies.

Sequencing also confirmed that, at both STR loci on chromosome 3, the child's allele matched one of the alleged father's but none of the mother's (Table 1). This was particularly significant for D3S1358, where the sequence variation supported the hypothesis of UPD rather than maternal mutation. Multiple mutations would otherwise be required to explain the observed alleles in the mother and child due to the variation in repeat numbers across two separate repeat units.

SNP data analysis

3.3

Genotypes were successfully obtained for all individuals at 273 SNP sites on chromosome 3. Among these, 43 mismatches were identified between the mother and child, while, in contrast, no mismatches were observed between the child and the alleged father (Table 2).

Furthermore, no mismatches between the mother and child were observed on any chromosome other than chromosome 3. However, a Mendelian inconsistency was detected between the child and the alleged father at rs1800624 on chromosome 6, which is likely attributable to a de novo mutation [24].

Consistent with these observations, the child exhibited complete homozygosity across all SNP loci on chromosome 3, whereas heterozygosity was observed across all other chromosomes These findings provide strong evidence for isodisomic UPD.

Kinship testing scenarios

3.4

To investigate the effect of UPD on kinship testing, STR panels of increasing size were used: Identifiler (15 STRs), VeriFiler (23 STRs), and a combined VeriFiler + SureID 27comp panel (42 STRs). For the Identifiler and VeriFiler kits, the single mismatch between the mother and child at D3S1358 was treated as a mutation, in line with the FSR guidance [11]. CLRs for each scenario are summarized in Table 3. As outlined in the methods, acceptance thresholds were 10,000 CLR for parentage and 100 CLR for secondary relationship; values below these thresholds were considered inconclusive.

Paternity was conclusively established in all cases, although the 15 STR panel produced a borderline inconclusive result in the father versus uncle scenario. Maternal testing, however, proved far more vulnerable to the effects of UPD, particularly in duo testing (mother–child only). In the mother versus aunt scenarios, all panels produced inconclusive results, and with 42 STRs the outcome falsely supported the aunt hypothesis, that is, an incorrect relationship determination.

Although increasing the number of STR loci is generally expected to enhance discriminatory power, in this case no false results were generated using the 15‐ and 23‐locus panels. This may be because these panels each included only one locus on chromosome 3, and the child's homozygous allele differed by just one repeat from the mother's, allowing the inconsistency to be masked as a mutation. By contrast, when multiple additional loci from the same chromosome were included, the accumulation of discrepancies resulted in false conclusions.

These findings align with those of Ma et al. [9], who reported reduced CLRs in UPD cases despite using expanded STR panels, although their study did not extend to secondary relationships. In the present case, however, the risk of false conclusions in secondary relationship testing is particularly concerning, as the inclusion of additional loci magnified the impact of Mendelian inconsistencies concentrated on a single chromosome.

DISCUSSION

4

Due to the rarity of UPD, this study could only assess its impact on kinship testing through a single identified case. It is therefore important to supplement these findings with evidence from the literature and to continue reporting new cases and their implications for forensic interpretations [25, 26].

One recommendation from previous research is to restrict exclusions in parentage testing to cases where inconsistencies occur on more than one chromosome [27]. Furthermore, as Mendelian inheritance typically holds true for chromosomes not affected by UPD, some authors have suggested excluding loci from the suspected UPD chromosome during kinship analysis [9]. While this approach may improve the validity of results, it requires prior confirmation of UPD, ideally through expanded STR typing, STR sequence analysis via NGS, or SNP microarray testing focused on the chromosome of interest.

Given the findings in this study, it is advisable to suspect UPD when two or more loci with inconsistencies are located on the same chromosome, with no mismatches elsewhere. It is also important to note that not all UPD cases involve isodisomy; in heterodisomy, mismatches may occur as isolated events and risk being misinterpreted as mutations.

The prevalence of UPD varies between chromosomes. Chromosomes 4, 15, and 16 are most commonly implicated, while chromosomes 18 and 19 are least frequently reported. Other chromosomes are associated with UPD less often. Notably, maternal UPD occurs at least twice as often as paternal UPD [28]. This disparity may be linked to epigenetic imprinting differences: maternal imprinting tends to support fetal development and placental maintenance, whereas paternal imprinting is more often associated with enhanced placental growth at the expense of fetal viability. Consequently, paternal UPD is more likely to result in spontaneous miscarriage or pregnancy loss [29]. In addition, UPD has been associated with advanced maternal and paternal age at conception, likely due to declining gamete production and functionality [10].

Children with UPD frequently present with clinical syndromes such as Prader–Willi, Angelman, Silver–Russell, and Beckwith–Wiedemann [30, 31, 32, 33]. These arise from loss of parental‐specific imprinting at key genes. However, not all individuals with UPD display clinical symptoms. As such, kinship testing may be requested in cases where the affected individual is unaware of any chromosomal abnormality.

In this case, paternal UPD on chromosome 3 was identified, an especially rare occurrence. Previous research indicates that this chromosome lacks known imprinting regions associated with pathological phenotypes, and no developmental anomalies were observed in the individual. The presence of UPD was suspected only due to inconsistencies detected during routine forensic testing; no prior medical history or clinical indicators suggested a chromosomal anomaly.

CONCLUSION

5

This study provides clear evidence of complete paternal isodisomy of chromosome 3, confirmed through conventional STR typing, STR sequencing by NGS, and SNP microarray genotyping. The findings demonstrate the challenges that UPD can present in forensic kinship testing, particularly in maternal and secondary relationship assessments, where inconclusive or misleading results may arise.

While paternity was reliably established across all tested scenarios, maternity and alternative relationship analyses were significantly affected, with expanded STR panels exacerbating the risk of false exclusions. These results emphasize that increasing the number of STR loci does not always enhance reliability in the presence of UPD.

To minimize the likelihood of erroneous conclusions, forensic practitioners must be aware of the genetic features of UPD and its potential impact on relationship testing. Incorporating automated software alerts to flag multiple inconsistencies restricted to a single chromosome would provide an important safeguard. Until such measures are routinely available, heightened analyst awareness and careful manual review remain essential.

By drawing attention to the forensic implications of UPD, this study highlights the need for improved recognition, reporting, and methodological adaptation to ensure more accurate and reliable outcomes in kinship assessments.

CONFLICT OF INTEREST STATEMENT

The authors declare no conflicts of interest.

ETHICS STATEMENT

This study was conducted with favorable ethical approval from the Human Ethics Committee at the University of Lincoln (reference: UoL2023_12459). Informed consent was obtained from all donors for the use of casework samples in method development and validation. All samples were anonymized, and no personal information or identifiable individual genotypes are published in this study.