Clinical application value of targeted amplicon sequencing technology in fetuses with uniparental disomy-related imprinting disorders: a multicenter study

Abstract

Background

To evaluate the application of targeted amplicon sequencing (TA-seq)—a method based on multiplex PCR and high-throughput sequencing—for the prenatal detection of uniparental disomy (UPD)-related imprinting disorders (ImpDis).

Methods

This retrospective study included 370 samples suspected of UPD from 42 hospitals across China. Of these, 294 samples were successfully tested by TA-seq, and methylation multiplex ligation-dependent probe amplification (MS-MLPA) being regarded as the reference method for methylation-based detection of imprinting disorders.

Results

TA-seq identified 36 positives and 258 negatives, of which 30 positives and 255 negatives were consistent with the findings from MS-MLPA. The sensitivity, specificity, positive predictive value, and negative predictive value of TA-seq were 90.9% (30/33), 97.7% (255/261), 83.3% (30/36), and 98.8% (255/258), respectively. The concordance between the two methods was 96.9% (285/294). Additionally, we observed that the positive rates vary widely among different testing indication groups. The group of ≥ 5 Mb region of homozygosity (ROH) detected by SNP-array on chromosomes 6, 7, 11, 14, 15, or 20 exhibited positive rates of 13.4% (17/127) and 11.0% (14/127) for TA-seq and MS-MLPA, respectively. In contrast, the group of familial or de novo balanced Robertsonian translocation or isochromosome involving chromosome 14 or 15 based on CVS or amniocentesis and the group of de novo small supernumerary marker chromosomes with no apparent euchromatic material in the fetus both demonstrated positive rates of 0% (0/23 and 0/6 for TA-seq and MS-MLPA, respectively).

Conclusions

This retrospective study demonstrates TA-seq as a potentially effective method for prenatal screening of UPD-related ImpDis. Its cost-effectiveness, adaptability, and reliability make it promising for future clinical use. However, limitations inherent in this study—including its retrospective design, lack of trio validation, and chromosomal bias among true positives—warrant further investigation before broader application of TA-seq.

Supplementary Information

The online version contains supplementary material available at 10.1186/s12967-025-07329-x.

Introduction

Imprinted genes are regulated by epigenetic mechanisms and exhibit monoallelic expression in a parent-of-origin-dependent manner. Imprinting disorders (ImpDis) arise when both parental alleles are either aberrantly expressed or silenced. Currently, 13 distinct ImpDis have been identified, which are associated with chromosomes 6, 7, 11, 14, 15, and 20 [1]. There are four types of molecular alterations in ImpDis, including uniparental disomy (UPD), copy number variation (CNV), single nucleotide variation (SNV), and imprinting defects [2]. UPD is responsible for nine ImpDis, including transient neonatal diabetes mellitus, Silver–Russell syndrome, Beckwith–Wiedemann syndrome, Temple syndrome, Kagami–Ogata syndrome, Prader–Willi syndrome, Angelman syndrome, pseudohypoparathyroidism type 1B, and Mulchandani–Bhoj–Conlin syndrome [3].

UPD refers to the situation in which both homologous chromosomes are inherited exclusively from one parent, presenting as three main types: (1) uniparental heterodisomy (UPhD), where a pair of homologous chromosomes originates from one parent; (2) uniparental isodisomy (UPiD), where two identical copies of a single chromosome originate from one parent; and (3) mixed UPD, which includes both UPhD and UPiD parts [3]. UPD is not clinically rare, with an overall prevalence of approximately 1 in 2,000 births in the general population [4]. Nevertheless, only a small proportion of UPDs are pathogenic and result in ImpDis [1, 5].

ImpDis significantly impact patient survival and overall quality of life. Prenatal screening and diagnosis can reduce the risk of these conditions. In 2001, the American College of Medical Genetics and Genomics (ACMG) released the first statement of diagnostic testing for UPD [6]. Then in 2020, the ACMG updated this statement to broaden the indications for UPD testing and clarify the distinctions between prenatal and postnatal applications [7]. China has also issued several consensus statements and guidelines in recent years, underscoring the importance of prenatal testing for UPD and UPD-related ImpDis [8–11].

Current prenatal testing for UPD-related imprinting disorders employs multiple approaches, including short tandem repeat analysis (STR), single nucleotide polymorphism array (SNP-array), exome sequencing (ES), targeted amplicon sequencing (TA-seq), and methylation-specific multiplex ligation-dependent probe amplification (MS-MLPA), each with distinct clinical applicability and cost profiles [2, 10]. TA-seq involves multiplex PCR amplification of selected genomic regions, followed by high-throughput sequencing. This method is robust, highly sensitive, and cost-effective [12], and has been widely applied in areas such as pathogenic microbe detection, tumor genomics, and genetic disease diagnostics [13, 14]. In this study, we developed a new prenatal detection method for nine UPD-related ImpDis based on TA-seq technology and assessed its detection performance through a retrospective multicenter study.

Materials and methods

Participant samples and clinical data were collected with informed consent and ethical approval from the Medical Ethics Committee of the First Affiliated Hospital of Zhengzhou University (Ethics Review Number: 2023-KY-0905-003).

Subjects

We collected 370 singleton pregnancy samples with varied indications for UPD testing from 42 Chinese hospitals, including chorionic villi, amniotic fluid, umbilical cord blood, and products of conception. Inclusion criteria followed the UPD testing statement published by ACMG [7], with minor modifications. Exclusion criteria included: (1) twin or multiple pregnancies; (2) ImpDis caused by CNV; (3) consanguineous parents; (4) triploidy.

Genomic DNA extraction and quality control

Genomic DNA was extracted using the MagPure DNA Micro Kit (Magen, China), following the manufacturer’s instructions. The quality and concentration of the genomic DNA were assessed using agarose gel electrophoresis and a Qubit 3.0 fluorometer, respectively. To exclude maternal cell contamination, we conducted STR analysis utilizing the Microreader 21 Direct ID System (Microreader Genetics, China). PCR products were analyzed on an ABI 3130XL sequencer (Applied Biosystems, USA), and genotypes were scored with GeneMapper 4.0 (Applied Biosystems, USA).

TA-seq

We designed multiplex PCR primers targeting 1,230 SNP loci across imprinted regions on chromosomes 6, 7, 11, 14, 15, and 20. These loci were selected from multiple polymorphic SNP databases (dbSNP, gnomAD, ExAC, and 1000 Genomes) (Fig. 1). The multiplex PCR reaction was performed in a 20 µL reaction system, including 5 µL of genomic DNA (4 ng/µL), 2 µL of M-primer (containing an index), 3 µL of UPD primer pool, and 10 µL of 2× multiplex PCR mix. The amplification conditions were as follows: 95 °C for 2 min; 20 cycles of 95 °C for 30 s and 60 °C for 4 min; finally 72 °C for 5 min; 4℃ forever. Purified PCR products were used to construct DNA libraries for sequencing. The libraries were sequenced in single-end mode at 40 bp using the Nextseq 550AR sequencer (Annoroad, China), yielding approximately 1,000,000 raw reads per sample. After sequencing, we used Cutadapt (version 1.10) to remove adapters and low-quality sequences using, and then mapped reads to the human reference genome (GRCh37/hg19) using Burrows-Wheeler Aligner (version 0.7.15) with the mem algorithm. VarScan (version 2.4.3) was utilized for SNV calling.

Fig. 1.

Circos plot displaying the imprinted regions and SNP loci on chromosomes 6, 7, 11, 14, 15, and 20. Fuchsia, red, and green bars represent centromeric regions, imprinted regions, and SNP loci of the imprinted genes, respectively. Blue and orange fonts represent paternally expressed genes and maternally expressed genes, respectively

Detected variants within the target area were classified empirically based on the variant allele frequency (VAF). A VAF within 0.5 ± 0.25 was considered heterozygous; otherwise, a homozygous mutation or wild type was considered. For each UPD target, a LOD score was derived by log transformation of the binomial probability density of the number of heterozygous loci within each targeted area,

k is the number of heterozygous loci, n is the expected total loci within each target area, and x is the Bernoulli probability, which under this scenario is assumed to be 50% for a general population. A LOD score above 20 is considered positive for UPiD, which is approximately 30 heterozygous sites out of 30 total loci.

To predict the target UPhD status of a test sample, it is essential to use samples from its biological parents. First, the number of identical genotypes between test sample and each parental donor is counted within the targeted area, denoted as X. Then, the number of common loci, irrespective of their genotype differences, is counted for each test sample and one parental sample, denoted as T. If the ratio of X/T is ≥ 95%, UPhD of this target in the test sample is suspected to have originated from the compared parental sample. The parental origin of positive UPiD was also similarly determined by calculating the genotype matching rate between test sample and individual biological parent; if the ratio of a test pair is above 95%, the parental origin is assigned accordingly.

MS-MLPA

MS-MLPA is regarded as the reference method for methylation-based detection of imprinting disorders. MS-MLPA analysis was performed using the SALSA MS-MLPA Probemix ME034-C1 multi-locus imprinting kit (MRC-Holland, The Netherlands) according to the manufacturer’s protocol. To ensure the stability and accuracy of the results, all MS-MLPA experiments were uniformly conducted in the Genetic and Prenatal Diagnosis Center Laboratory of the First Affiliated Hospital of Zhengzhou University, with a single experimenter handling both execution and analysis. Amplified products were analyzed using the ABI 3130XL sequencer (Applied Biosystems, USA) with the GeneMapper 4.0 (Applied Biosystems, USA). Peak area values were analyzed by Coffalyser V7 software (MRC-Holland, The Netherlands). The formula used for methylation dosage ratio (MR) was as follows:

where Px is the peak area of a given target probe, Pctrl is the sum of the peak areas of all control probes, Dig stands for HhaI digested sample, and Undig stands for undigested sample.

Statistical analysis

Statistical analyses, including positive detection rate, sensitivity, specificity, positive predictive value (PPV), negative predictive value (NPV), and concordance, were performed using SPSS version 26 (IBM SPSS, USA). Categorical variables appear as percentages, and continuous variables as mean ± standard deviation.

Results

Subjects

The study initially enrolled 370 samples from 42 hospitals. Then 76 samples were excluded due to CNV, triploidy, consanguineous parents, maternal cell contamination, or insufficient DNA concentration for experiments. The remaining 294 samples were successfully assayed using both TA-seq and MS-MLPA (Fig. 2). Maternal ages ranged from 19 to 45 years, with an average of 30.9 ± 5.2 years. Gestational weeks at sampling varied by procedure: chorionic villus sampling (9–13 weeks; mean 11.8 ± 1.1), amniocentesis (16–33 weeks; mean 20.4 ± 3.5), products of conception (6–26 weeks; mean 10.6 ± 4.3), and cordocentesis (26–36 weeks; mean 30.1 ± 2.5). Amniotic fluid comprised the majority of specimens (n = 209), followed by chorionic villi (n = 38), products of conception (n = 38), and umbilical cord blood (n = 9).

Fig. 2.

Flowchart of TA-seq and MS-MLPA results. CNV, copy number variation; MCC, maternal cell contamination; MS-MLPA, methylation multiplex ligation-dependent probe amplification; NPV, negative predictive value; PPV, positive predictive value; TA-seq, targeted amplicon sequencing

The six testing indication groups varied substantially in sample size. The largest cohort (n = 127) was the group defined by ‘≥ 5 Mb region of homozygosity (ROH) detected by SNP-array on chromosomes 6, 7, 11, 14, 15, or 20’ (GROUP-ROH), followed by 78 cases in the group with ‘trisomy or monosomy of chromosomes 6, 7, 11, 14, 15, or 20 detected by non-invasive prenatal screening but disomy or mosaicism by prenatal diagnosis’ (GROUP-NIPS). The remaining four indication groups comprised GROUP-MOSAIC (≥ 10% mosaicism for trisomy or monosomy detected by prenatal diagnosis; n = 43), GROUP-RT (familial or de novo balanced Robertsonian translocation or isochromosome involving chromosome 14 or 15 based on CVS or amniocentesis; n = 23), GROUP-CHM (whole-genome paternal UPD detected by SNP-array; n = 17), and GROUP-sSMC (de novo small supernumerary marker chromosomes with no apparent euchromatic material in the fetus; n = 6).

MS-MLPA results

MS-MLPA analysis of 294 samples revealed no microdeletions within the probe range, confirming that observed methylation alterations were unrelated to microdeletions. MS-MLPA detected 33 cases with aberrant methylation, including 1 case with loss of methylation (LoM) of PLAG1:TSS-DMR, 7 with gain of methylation (GoM) of PLAG1:TSS-DMR, 1 with GoM of MEG3:TSS-DMR and LoM of MEG8:int2-DMR, 2 with LoM of MEG3:TSS-DMR and GoM of MEG8:int2-DMR, 5 with GoM of SNRPN, and 17 with GoM of all paternal imprinted alleles and LoM of all maternal imprinted alleles (Supplementary Table 1).

TA-seq performance

TA-seq yielded a positive detection rate of 12.2% (36/294). Compared to MS-MLPA, TA-seq detected 30 true positives, 6 false positives, 3 false negatives, and 255 true negatives. Its sensitivity, specificity, PPV, and NPV were 90.9% (30/33), 97.7% (255/261), 83.3% (30/36), and 98.8% (255/258), respectively. The concordance between the two methods reached 96.9% (285/294).

False negatives and false positives

Three cases with UPD-related ImpDis were not detected by TA-seq, representing false negatives (Table 1; Supplementary Fig. 1A-1 C). One case was referred with a testing indication of ‘trisomy 15 detected by NIPS but disomy 15 by CNV-seq,’ while the other two were referred with indications of ‘25.7 Mb ROH on chromosome 15 detected by SNP-array at 15q14q22.2’ and ‘37.3 Mb ROH on chromosome 14 detected by SNP-array at 14q11.2q23.1,’ respectively.

Table 1.

Summary of 3 false negatives and 6 false positives ^a

	TA-seq	MS-MLPA	Indication group	Specific indications
1	Not detected	GoM of SNRPN	GROUP-NIPS	trisomy 15 detected by NIPS but disomy 15 by CNV-seq
2	Not detected	GoM of SNRPN	GROUP-ROH	25.7 Mb ROH on chromosome 15 detected by SNP-array at 15q14q22.2
3	Not detected	GoM of MEG3:TSS-DMR and LoM of MEG8:int2-DMR	GROUP-ROH	37.3 Mb ROH on chromosome 14 detected by SNP-array at 14q11.2q23.1
4	UPD(6)	No methylation alteration	GROUP-ROH	7.2 Mb ROH detected by SNP-array at 6q24.1q24.3
5	UPD(6)	No methylation alteration	GROUP-ROH	5.7 Mb ROH detected by SNP-array at 6q23.3q24.2.
6	UPD(6)	No methylation alteration	GROUP-ROH	11 Mb ROH detected by SNP-array at 6q23.3q25.1
7	UPD(7)	No methylation alteration	GROUP-ROH	25.6 Mb ROH detected by SNP-array at 7q22.3q32.3
8	UPD(11)	No methylation alteration	GROUP-NIPS	trisomy 7 was detected by NIPS but disomy 7 by CNV-seq
9	UPD(20)	No methylation alteration	GROUP-ROH	10.4 Mb ROH detected by SNP-array at 20q13.2q13.33

^aAbbreviations: LoM, loss of methylation; GoM, gain of methylation. CNV-seq, copy number variation sequencing; ROH, region of homozygosity; DMR, differentially methylation regions

Six normal cases were incorrectly identified as positive by TA-seq, representing false positives (Table 1; Supplementary Fig. 1D-1I). Five cases exhibited large ROHs identified by SNP-array, with sizes ranging from 5.7 to 11 Mb. The remaining case was referred with a testing indication of ‘trisomy 7 detected by NIPS but disomy 7 by CNV-seq’.

Positive detection rates of TA-seq and MS-MLPA under 6 testing indication groups

Although sample sizes varied substantially across indication groups, TA-seq and MS-MLPA produced broadly consistent positive detection rates (Fig. 3). In the GROUP-ROH cohort, TA-seq detected 13.4% (17/127) of cases as positive, compared with 11.0% (14/127) by MS-MLPA. Both methods identified all GROUP-CHM cases as positive (17/17, 100%). In GROUP-MOSAIC, each method detected 2.3% (1/43) of cases as positive, while in GROUP-NIPS, both identified 1.3% (1/78) of cases as positive. Neither TA-seq nor MS-MLPA detected any positive cases in GROUP-RT or GROUP-sSMC.

Fig. 3.

Bar graphs were plotted in Excel to display the positive detection rates in GROUP-ROH, GROUP-MOSAIC, and GROUP-NIPS as detected by TA-seq and MS-MLPA. The numerator in parentheses denotes the number of positive samples detected for each chromosome within an indication group by either TA-seq or MS-MLPA, whereas the denominator represents the total number of samples available for that chromosome in the same group

Discussion

In this multicenter retrospective study, we evaluated a novel TA-seq technique for the prenatal detection of UPD-related ImpDis. TA-seq exhibits high sensitivity, specificity, PPV, and NPV, indicating its high detection efficiency, which approaches that of MS-MLPA.

Causes of false negatives and false positives

TA-seq identified three false-negative cases. The underlying reasons for two of these missed detections were easily explained: SNP-array analysis revealed a large ROH in each case (14q11.2q23.1 and 15q14q22.2, respectively), indicating a “mixed UPD” type. Both ROHs fell outside the amplicon regions of TA-seq, which resulted in negative TA-seq calls due to retained heterozygosity within those amplicons. The third case lacked a definitive explanation due to unavailable SNP-array data. We speculate that the reason may be analogous to the other two cases, suggesting its UPD type could also be a “mixed UPD” in which ROH fell outside the amplicons. Alternatively, we cannot exclude the possibility of UPhD. Indeed, all DNA sequencing techniques—such as STR analysis, SNP-array, and ES—fail to identify UPhD when applied to single samples [10]. Hoppman et al. [15] reported that 33% of UPhD cases could be missed by a single-sample SNP-array due to the lack of extended ROH. Since TA-seq targets imprinted regions rather than the entire chromosome, single-sample TA-seq may increase its risk of missed detections relative to other techniques. The strategy for solving this problem is to use trio-based analysis [16, 17].

Trio-based analysis also offers further advantages. Paternal UPD(6) and maternal UPD(7) demonstrate clear pathogenicity, but their reciprocal patterns—maternal UPD(6) and paternal UPD(7)—are not associated with clinical phenotypes [3]. Hence, trio-based analysis can ascertain disease presence. For UPD(11), UPD(14), UPD(15) and UPD(20), both paternal and maternal origins exhibit pathogenic effects but lead to distinct ImpDis. For example, paternal UPD(11) causes Silver–Russell syndrome, while maternal UPD(11) results in Beckwith–Wiedemann syndrome; paternal UPD(15) causes Angelman syndrome, whereas maternal UPD(15) results in Prader–Willi syndrome. Thus, trio-based analysis improves diagnostic accuracy by determining both the presence and specific type of ImpDis, thereby facilitating comprehensive genetic counseling and supporting early postnatal management and treatment. However, parental samples were unavailable in this retrospective study. Although we have confidence that trio-based TA-seq analysis can avoid false negatives, its clinical performance requires further validation in a multicenter prospective study.

The number of false positives is twice that of false negatives, primarily due to the limited number of true positive samples. We therefore adopted a slightly lenient threshold for positive calls to reduce the risk of missed detection. However, false positives increase the workload for subsequent diagnosis and genetic counseling, along with parental psychological stress. Future efforts should prioritize expanding true-positive sample collections while concurrently optimizing bioinformatic algorithms or refining the threshold to improve TA-seq accuracy.

Differences in positive detection rates under different testing indication groups

Positive detection rates varied substantially (0-100%) across the six testing indication groups. Both GROUP-CHM and GROUP-ROH exhibit high positive rates, suggesting that a large ROH serves as a more optimal indication than others. Within GROUP-ROH, chromosome 6, 14, or 15 had higher true positive rates than chromosome 7, 11, or 20. We propose that the observed true positive bias likely stems from the following factors: (1) substantial variation in UPD incidence across chromosomes, with notably high rates for UPD(14) and UPD(15) but minimal occurrence for UPD(7) and UPD(11) [18]; (2) potential selection bias during sample collection, influenced by hospital retention practices, sample sharing protocols, and quality control filtering; (3) the limited cohort size, as only 294 cases were analyzed using both TA-seq and MS-MLPA; and (4) the presence of polymorphic ROHs. Previous studies have identified polymorphic ROHs at 7q11.22q11.23 [19], 11p11.2p11.12 [19, 20], and 20q11.21q11.23 [19, 21]. In our study, the same ROHs were also observed across multiple samples in GROUP-ROH, with frequencies of 57.1% (24 cases) for 7q11.22q11.23, 24% (6 cases) for 11p11.2p11.12, and 60% (9 cases) for 20q11.21q11.23. None of these cases exhibited methylation anomalies in MS-MLPA analysis, suggesting that ROHs confined to these chromosomal regions are not recommended as UPD testing indications.

Recent studies have reported that a positive NIPS result for aneuploidy serves as a testing indication for UPD [22–24]. For instance, Ngo et al. [18] identified two UPD(9), one UPD(15), and two UPD(16) from 35 high-risk pregnancies; In a larger cohort, Hu et al. [25] found 30 fetuses with UPDs from 528 gravidas, including five cases of UPD(6), five cases of UPD(7), plus single cases of UPD(11), UPD(14), and UPD(15). Lannoo et al. [26] summarized multiple cases with rare autosomal trisomies found by NIPT in order to assess their risks for UPD. No significant correlation was observed between trisomy 6 and paternal UPD(6). A weak correlation was observed between trisomy 7 and maternal UPD(7), trisomy 14 and UPD(14), and trisomy 20 and UPD(20). Trisomy 15 indicates a high risk of maternal UPD(15) but a very low risk of paternal UPD(15). In the present study, we enrolled 78 high-risk NIPS cases, comprising 3, 44, 6, 9, 11, and 5 cases of trisomy 6, 7, 11, 14, 15, and 20, respectively. MS-MLPA detected only one case with a methylation anomaly on chromosome 15 (gain of methylation of SNRPN). Given the testing indication, this imprinting anomaly probably results from maternal UPD(15). Despite the limited sample size, our results seem to reflect the potential influence of chromosome 15 aneuploidy on UPD.

Advantages and disadvantages of TA-seq

STR analysis, SNP-array, and ES are currently-used methods for detecting UPD-related ImpDis. Compared to these three methods, TA-seq is more cost-effective than SNP-array and ES for UPD detection (Table 2). TA-seq also demonstrates greater versatility by aligning directly with NIPS or CNV-seq workflows, substantially streamlining experimental procedures. In contrast, STR analysis relies on first-generation sequencing platforms, creating additional spatial and financial challenges for laboratories equipped with next-generation sequencing platforms. However, TA-seq has limited sensitivity in detecting mixed UPD since it can only detect cases where ROHs occur around imprinted regions (Table 2). By comparison, SNP-array and ES can effectively detect mixed UPD regardless of its ROH location, owing to their comprehensive probe coverage across entire chromosomes.

Table 2.

A comparison between the methods in detecting UPD-related imprinting disorders

Technique	UPhD	UPiD	Mixed UPD	Cost
TA-seq	——	YES	Limited	low
STR analysis	——	YES	Limited	low
SNP-array	——	YES	YES	high
ES	——	YES	YES	high

Limitations of this study

There are also limitations in this study. First, we did not validate trio-based TA-seq using prenatal samples, and second, the number of true positive cases remains limited. Moreover, the true positives are concentrated on chromosomes 6, 14, and 15. To address these limitations, we are planning a multicenter prospective study enrolling trios of prenatal samples to comprehensively evaluate the detection performance of TA-seq.

Supplementary Information

Below is the link to the electronic supplementary material.

Abbreviations

ImpDis

Imprinting disorders

CNV

Copy number variation

SNV

Single nucleotide variation

UPD

Uniparental disomy

UPhD

Uniparental heterodisomy

UPiD

Uniparental isodisomy

ACMG

American College of Medical Genetics and Genomics

TA-seq

Targeted amplicon sequencing

MS-MLPA

Methylation-specific multiplex ligation-dependent probe amplification

STR

Short tandem repeat

SNP-array

Single nucleotide polymorphism array

ES

Exome sequencing

NIPS

Non-invasive prenatal screening

ROH

Regions of homozygosity

GROUP-ROH

The group with ‘≥ 5 Mb region of homozygosity (ROH) detected by SNP-array on chromosomes 6,7,11,14,15,or 20’

GROUP-NIPS

The group with ‘trisomy or monosomy of chromosomes 6,7,11,14,15,or 20 detected by non-invasive prenatal screening, but disomy or mosaicism by prenatal diagnosis’

GROUP-MOSAIC

The group with ‘≥ 10% mosaicism for trisomy or monosomy detected by prenatal diagnosis’

GROUP-RT

The group with ‘familial or de novo balanced Robertsonian translocation or isochromosome involving chromosome 14 or 15 based on CVS or amniocentesis’

GROUP-CHM

The group with ‘whole-genome paternal UPD detected by SNP-array’

GROUP-sSMC

The group with ‘de novo small supernumerary marker chromosomes with no apparent euchromatic material in the fetus’ group

Author contributions

Ning Liu and Xiangdong Kong conceived and designed the study, wrote the study protocol, and contributed to the acquisition of clinical data. Lina Liu, Yin Feng and Panlai Shi performed the experiments, statistical analyses. Authors from collaborating institutions provide qualified samples and clinical information. All of the authors reviewed and commented on the manuscript and approved the final version.

Funding

This study was supported by Special Fund for Key Research, Development and Promotion of Science and Technology of Henan Province (222102520018).

Data availability

The datasets used in the current study are available from the corresponding author on reasonable request.

Declarations

Ethics approval and consent to participate

Donors and patient provided written informed consent.

Consent for publication

All authors consent for publication.

Competing interests

The authors declare that they have no competing interests.