Borderline Z-scores in non-invasive prenatal screening: does its presence hold clinical significance?

Abstract

Background

This study aimed to investigate the utility of repeated testing for the detection of fetal cell-free DNA (cfDNA) from maternal peripheral blood in cases with borderline Z-scores and to analyze the associated pregnancy outcomes.

Methods

A retrospective analysis was conducted on 83,443 pregnant women who voluntarily underwent non-invasive prenatal testing (NIPT) at the Affiliated Women and Children’s Hospital of Ningbo University between January 2020 and January 2024. Pregnant women whose initial NIPT results indicated borderline Z-scores were subsequently followed up.

Results

Among 83,443 pregnant women undergoing NIPT, 700 cases (0.84%) initially showed borderline Z-scores. After retesting, this number decreased to 211 cases (0.25%) and further decreased to 26 cases (0.03%) after re-sampling. Among the initial 700 cases, 34 exhibited abnormal NIPT results, corresponding to a positive rate of 5.29%. Subsequent prenatal diagnosis confirmed a total of six fetal abnormalities, including one case of trisomy 13 mosaicism, one case of trisomy 21 mosaicism, one case of 47,XXY, two cases of copy number variations (CNVs), and one case with B-ultrasound findings indicating an abnormality (nuchal translucency [NT] of 3.5 mm and omphalocele). The overall rate of fetal abnormalities was 0.9%.

Conclusion

Initial NIPT outcomes suggested elevated screening positive rates and a higher incidence of fetal abnormalities among cases with borderline Z-scores compared to the general population. Establishing a defined borderline Z-score threshold in NIPT protocols is crucial to mitigate the risk of missed screenings. Implementing re-construction and/or re-sampling procedures significantly reduces the failure rate attributed to borderline Z-scores, facilitating the accurate identification of most of pregnancies with normal fetal development and decreasing the need for unnecessary invasive prenatal diagnostic interventions. Pregnant women with detection failures due to borderline Z-scores should be actively counseled and encouraged to pursue prenatal diagnosis.

Introduction

Non-invasive prenatal testing (NIPT) is a screening method that analyzes cell-free fetal DNA (cffDNA) in the maternal peripheral blood through high-throughput sequencing. The sequencing data are compared with a model of normal pregnant women’s data, and risk assessment is performed to determine the risk of fetal chromosomal abnormalities. NIPT has high specificity and sensitivity for fetal chromosomal diseases, especially for trisomy 21, trisomy 18, and trisomy 13, making it an important part of prenatal screening [1–3]. For fetal chromosomal aneuploidies, results are typically expressed as Z-scores, indicating whether there is a statistically significant difference between the tested sample and normal pregnant women. The Z-score is a risk judgment value obtained after processing the sequencing data with GC correction and other treatments [4]. The normal Z-score range is − 3 to + 3. When the Z-score is ≥ 3, it suggests a high risk of chromosomal aneuploidies in the fetus, and the higher the Z-score, the higher the positive predictive value (PPV) [5].

In practical application, due to the inherent variability in experimental data, different NIPT technology platforms set a borderline Z-score threshold, also known as the “gray zone” for Z-scores, it cannot be definitively classified as high risk or low risk, leading to uncertainty results. If a borderline Z-score is not established, some normal fetuses may be misclassified as high risk, resulting in an increased false-positive rate and unnecessary invasive prenatal diagnostic procedures. Conversely, some abnormal fetuses may be missed, leading to false-negative results. Research indicates that the borderline Z-score is a primary cause of NIPT experimental failure, and repeated testing can effectively reduce the false-positive rate associated with the borderline Z-score [6, 7]. This study conducted a retrospective analysis of samples that fell within the borderline Z-score from January 2020 to January 2024, followed up their pregnancy outcomes, and discussed the necessity of setting a Z-score in the detection of cell-free fetal DNA in peripheral blood and the value of repeat testing.

Methods

Study subjects

Data were collected from 83,443 pregnant women who underwent NIPT at The Affiliated Women and Children’s Hospital of Ningbo University from January 2020 to January 2024. Inclusion criteria: voluntarily accepted NIPT testing and were at a gestational age of 12–26⁺⁶ weeks. Exclusion criteria: received allogeneic blood transfusion, transplantation surgery, etc., within the past year; received immunoglobulin treatment within 4 weeks; triplets or more; ultrasound indicated fetal structural abnormalities; had a clear history of chromosomal abnormalities or a family history of genetic diseases; coexistence with malignant tumors. For pregnant women at high risk of NIPT, invasive prenatal diagnosis was performed after informed consent was obtained.

All pregnant women signed an informed consent form. This study has been reviewed and approved by the hospital’s ethics committee (2024KYSL-068).

Research methods

NIPT testing

Peripheral blood samples of 10 mL were collected from pregnant women using K- tubes, and plasma was separated within 72 h. The extraction and construction of cell-free DNA libraries were performed using the Fetal Chromosomal Aneuploidy (T21, T18, T13) Detection Kit (based on the Combinatorial Probe-Anchor Synthesis Sequencing Method) and the MGISP-960 High-Throughput Automated Nucleic Acid Extraction and Library Preparation System. High-throughput sequencing was conducted using the Sequencing Reaction Universal Reagent Kit (based on the Combinatorial Probe-Anchor Synthesis Sequencing Method) and the MGISEQ-2000 sequencing platform. All reagents and instruments were procured from BGI Genomics Co., Ltd., Shenzhen, China. Following sequencing, the generated data were analyzed to determine the Z-scores for chromosomes 21, 18, and 13. A Z-score within the range of − 3 to + 3 indicates a low risk, while a Z-score of ≥ 3 suggests a high risk. Any other abnormalities were reported in an additional supplementary report format.

In this study, the BGI HALOS Informatics Analysis System was employed, which has established a borderline Z-score based on accumulated experimental data. When the NIPT results showed a Z-score within the range of + 1.96 to + 3, the system flagged the sample as falling within the borderline Z-score, prompting a re-construction of the DNA library. If the re-constructed library results still indicated a borderline Z-score, the pregnant woman was notified via telephone to re-sample for further verification. For cases where both blood draws consistently indicated a borderline Z-score, invasive prenatal diagnostic procedures were recommended.

Prenatal diagnosis

For pregnant women at high risk based on NIPT results, amniocentesis is performed under ultrasound guidance after informed consent, and 35 ml of amniotic fluid is extracted for prenatal diagnosis (chromosomal karyotype analysis and/or CMA). Chromosomal karyotype analysis: amniotic fluid cells are cultured, harvested, prepared, and analyzed for karyotype after G-banding, with karyotype descriptions based on the International System for Human Cytogenetic Nomenclature (ISCN2016). CMA uses the whole-genome CytoScantm 750k chip (Affymetrix, USA), and the ChAS4.0 (Chromosome Analysis Suite 4.0, Affymetrix) software is used to calculate and analyze the scanning image results, referring to international public databases including DGV, ClinGen, and ClinVar, to assess the clinical significance of copy number variations.

Statistical methods

The data were statistically analyzed using SPSS 19.0 software. Count data are expressed as n, and measurement data are presented as mean ± standard. Correlation between variables was assessed using linear regression analysis.

Results

General situation of NIPT test failures

Among 83,443 pregnant women, the initial NIPT results indicated a borderline Z-score in 700 cases (0.84%, 700/83,443). Following library re-construction, 211 cases (0.25%, 211/83,443) remained within the borderline Z-score. After repeat blood sampling, 26 cases (0.03%, 26/83,443) still exhibited borderline Z-scores. Among the 700 cases requiring library re-construction, the most frequently observed borderline Z-scores were associated with T18 (37.9%, 265/700), followed by T21 (35.3%, 247/700) and T13 (22.7%, 159/700). Similarly, among the 206 cases requiring re-sampled, the most common borderline Z-scores were for T18 (45.1%, 93/206), followed by T21 (28.1%, 58/206) and T13 (18.4%, 38/206). Out of the 700 pregnant women, 34 cases (5.29%, 34/700) showed abnormal NIPT results. Further details are summarized in Table 1 and Fig. 1.

Table 1.

The distribution of borderline Z-score types among the 700 pregnant women after re-construction or re-sampling

Borderline Z-score types	Negative after re-construction	Positive after re-construction	Other chromosomal abnormalities after re-construction	Borderline after retest						Total
				Negative after re-sampled	Positive re-sampled	Other chromosomal abnormalities after re-sampled	Borderline after re-sampled	No re-sampled	Total
T13	117	3	0	29	1	1	7	1	39	159
T21	181	2	4	48	1	3	6	2	60	247
T18	163	4	3	72	8	2	11	2	95	265
T13, T18	2	0	0	5	0	0	1	0	6	8
T13, T21	6	0	0	1	0	0	0	0	1	7
T21, T18	2	0	0	5	1	0	1	0	7	9
T13, T21, T18	1	0	1	3	0	0	0	0	3	5
Total	472	9	8	163	11	6	26	5	211	700

Fig. 1.

The NIPT results of the 700 pregnant women with borderline Z-score after re-construction or re-sampling

The correlation between borderline Z-scores and fetal fraction (FF)

In the initial test of 700 cases, the mean FF was 11.04%, with the proportion of borderline Z-scores of 100%. After re-construction, the mean FF increased to 11.25%, and the proportion of borderline Z-scores decreased to 30.14% (211/700). Among the 206 cases that underwent re-sampled, the mean FF was 11.91%, with a proportion of borderline Z-scores of 12.62% (26/206). A negative correlation was observed between mean FF and the proportion of borderline Z-scores (R² = 0.659). The cases were further stratified based on the type of aneuploidy indicated by borderline Z-scores, namely T21, T18, and T13. For cases with T21 borderline, the mean FF from the three tests was 10.87%, 11.09%, and 12.08%, witha  proportion of borderline Z-scores of 100%, 24.29%, and 10.34%, and an R² of 0.555. In the T18 borderline, the mean FF was 10.50%, 10.71%, and 11.62%, with a proportion of borderline Z-scores of 100%, 35.85%, and 11.83%, and an R² of 0.678. For the T13 borderline, the mean FF was 11.98%, 12.15%, and 13.18%, with a proportion of borderline Z-scores of 100%, 24.53%, and 18.42%, and an R² of 0.437. Detailed results are presented in Table 2 and Figs. 2 and 3.

Table 2.

Correlation analysis between different borderline Z-score types and FF

Borderline Z-score types	First		Second		Third
	Mean FF (%)	Proportion of borderline (%)	Mean FF (%)	Proportion of borderline  (%)	Mean FF (%)	Proportion of borderline  (%)
Total	11.04 ± 4.64	100	11.25 ± 4.84	30.14	11.91 ± 5.39	12.62
T21	10.87 ± 5.21	100	11.09 ± 5.41	24.29	12.08 ± 7.14	10.34
T18	10.50 ± 3.76	100	10.71 ± 3.86	35.85	11.62 ± 4.37	11.83
T13	11.98 ± 4.86	100	12.15 ± 5.11	24.53	13.18 ± 5.37	18.42

Fig. 2.

The relationship between different borderline Z-score types and FF. 2A: total; 2B: borderline of T21; 2C: borderline of T18; 2D: borderline of T13

Fig. 3.

Correlation analysis between different borderline Z-score types and FF. 3A: total; 3B: borderline of T21; 3C: borderline of T18; 3D: borderline of T13

Prenatal diagnosis situation

Among the 700 pregnant women, 65 cases were identified by NIPT as high-risk or other chromosomal abnormalities, no re-sampled, and experienced NIPT detection failure due to borderline Z-scores. They were advised to seek prenatal diagnostic consultation. Among the 65 pregnant women, 31 underwent invasive prenatal diagnosis, 29 refused invasive prenatal diagnosis, and 5 were lost to follow-up. The prenatal diagnosis rates for cases with high-risk results, other chromosomal abnormalities, no re-sampled, and borderline Z-scores were 80% (16/20), 50% (7/14), 0% (0/5), and 23.1% (6/26), respectively (Table 3).

Table 3.

Prenatal diagnosis situation of 65 pregnant women

NIPT results	n	Underwent prenatal diagnosis				Did not underwent prenatal diagnosis			Loss to follow-up
		True positive	Other chromosomal abnormalities	False positive	Total	Induced abortion	Birth	Total
High-risk	20	2	1	13	16	1	3	4	0
Other chromosomal abnormalities	14	0	1	7	8	0	6	6	0
No re-sampled	5	0	0	0	0	0	4a	4	1
Borderline Z-scores	26	0	1	6	7	0	15	15	4
Total	65	2	3	26	31	1	28	29	5

^aIn one case of a twin pregnancy, selective reduction was performed due to prenatal diagnosis ultrasound indicating abnormalities in one fetus, while the other fetus was born

Prenatal diagnosis results

Among the pregnant women who underwent prenatal diagnostic consultation, five cases of chromosomal abnormalities were detected, including one case of trisomy 13 mosaicism, one case of trisomy 21 mosaicism, one case of Klinefelter syndrome, and two cases of copy number variation (CNV) abnormalities. Among these, one of the CNV abnormalities was classified as a variant of uncertain clinical significance (VUS), the pregnant woman chose to continue the pregnancy, and no abnormalities were observed postnatally. The remaining four pregnant women opted for induced abortion. Among the pregnant women who did not undergo prenatal diagnosis, one case with a borderline trisomy 18 at 16 weeks of gestation experienced an unexplained miscarriage, and one case with borderline trisomy 18 in a twin pregnancy chose fetal reduction due to abnormal findings on prenatal diagnostic ultrasound (NT 5.3 mm, omphalocele) in one of the fetuses. In summary, the fetal abnormality rate was 0.9% (6/700). Follow-up did not identify any false-negative cases. For details, see Table 4.

Table 4.

Prenatal diagnosis results of 6 pregnant women

Case	Age (years)	Gestational age (weeks)	Clinical indicators	Borderline Z-score types	First Z-score	Z-score after re-construction	NIPT results	Prenatal diagnosis results	Follow-up
1	27	18+4	Abnormal MOM	T13	2.727	3.000	T13	arr[hg19]15q21.1 (44,927,065–45,904,982) × 3	Birth
2	34	13+3	Voluntary request	T21	2.912	3.570	T21	46, XN, 1qh + , rob (21:21) [17]/46, XN, 1qh + [40]	Induced abortion
3	39	17	Maternal age > 35	T13	2.146	3.29	T13	47, X?,  + 13 [34]/46, X? [4]	Induced abortion
4	36	14+2	Maternal age > 35	T21	2.997	0.558	XXY	47, XXY	Induced abortion
5	23	21+4	Serological screening borderline risk	T21	2.877	2.048	Detection failure	arr [hg19] 3q29 (195, 703, 616–197, 356, 334) × 1	Induced abortion
6	34	15	IVF twins	T18	2.014	2.541	Detection failure	NT 5.3 mm, omphalocele	Fetal reduction

True-positive detection rates for different borderline Z-score types

Using invasive prenatal diagnosis and/or prenatal diagnostic ultrasound as the gold standard, the true-positive detection rates for different types of borderline Z-score were analyzed. The results showed that the true-positive rates for T13, T21, and T18 borderline Z-score were 1.3%, 1.2%, and 0.4%, respectively, with positive predictive values (PPVs) of 40%, 30%, and 6.7%, respectively. No true positives were detected in the other cases (Table 5).

Table 5.

True-positive detection rates for different borderline Z-score types

Borderline Z-score types	Number of individuals in the borderline zone	Number of individuals in the borderline zone after retest	Underwent prenatal diagnosis	True positive		PPV (%)
				n	%a
T13	159	38	5	2	1.3	40
T21	247	58	10	3	1.2	30
T18	265	93	15	1	0.4	6.7
Otherb	29	17	1	0	0	0
Total	700	206	31	6	0.9	19.4

PPV true positives/number of prenatal diagnostic procedures;

^aTrue-positive rate = true positives/number of critical zone individuals (initial)

^bIndividuals with two or more types of critical zone categories

Discussion

The National Health and Family Planning Commission issued the “Technical Specifications for Prenatal Screening and Diagnosis of Cell-Free Fetal DNA in Maternal Peripheral Blood,” which states that the failure rate of NIPT should not exceed 5% [8]. An increasing number of domestic and international scholars have conducted analyses and discussions on the reasons of NIPT failure [9, 10]. The NIPT testing process is lengthy and complex, susceptible to various influencing factors, including sample-related issues such as hemolysis and coagulation, as well as experimental procedural issues such as DNA extraction and library construction failures. Therefore, in practical clinical practice, NIPT testing often requires library re-construction and/or re-sampled to reduce the failure rate. The borderline Z-score is a common reason for library re-construction and/or re-sampled in NIPT.

The Technical Specifications for Prenatal Screening and Diagnosis of Cell-Free Fetal DNA in Maternal Peripheral Blood defines the serum screening risk value between the high-risk cutoff and 1/1000 as the borderline risk [8]. The chromosomal abnormality rate of the serum screening borderline risk group is significantly higher than that of the low-risk group [11]. Therefore, the authors believe that a borderline Z-score may be established in NIPT. In this study, among the 700 pregnant women whose initial NIPT results indicated a borderline Z-score, 34 cases with abnormal results were detected, with a screening positive rate of 5.29% (34/700). Following prenatal diagnosis, a total of six fetal abnormalities were detected (including one case with prenatal diagnosis ultrasound indicating abnormalities and five cases with invasive prenatal diagnosis indicating chromosomal abnormalities), with a fetal abnormality rate of 0.9% (6/700). Without the establishment of a borderline Z-score, these abnormal fetuses might have been missed. This underscores the necessity of setting a borderline Z-score in NIPT to reduce the rate of missed screenings. Pan SQ et al. [12] analyzed over 50,000 NIPT screening cases from this laboratory, reporting screening positive and fetal abnormal rates of 0.6% (326/54957) and 0.2% (135/54957), respectively, which are significantly lower than those observed in this study. Similarly, Xiang L et al. [1] and Liu J et al. [3] analyzed NIPT screening data from the general population, and their screening positive and fetal abnormal rates were significantly lower than those in this study. These findings indicate that the screening positive rate and fetal abnormality rate in cases with initial NIPT results indicating a borderline Z-score are higher than those in the general population, which is consistent with the studies by Luo YM et al. [6], Chan N et al. [13], and Lu Y et al. [14].

Qi H et al. [14] included 6110 pregnant women who underwent NIPT using the Illumina NextSeq 550 platform. A Z-score between 3 and 4 for the target chromosomes (13, 18, 21) was defined as the borderline Z-scores. Among them, 160 cases (2.62%) fell into borderline Z-scores. After re-construction and re-sampled, the rates were reduced to 0.20% (12/6,110) and 0%, respectively. Luo YM et al. [11] enrolled 40,311 pregnant women for NIPT using the MGISEQ-500 platform. A Z-score in the range 1.96 < Z < 4 was considered to indicate a borderline Z-score for the target chromosomes. They observed a borderline Z-score incidence of 1.64% (663/40,311). Following re-construction and re-sampled, the rates decreased to 0.23% (91/40,311) and 0.02% (7/40,311), respectively. This study analyzed NIPT-related data from over 80,000 pregnant women and identified 700 cases (0.84%, 700/83,443) with initial borderline Z-scores (Z-scores between 1.96 and 3), which is lower than the reported by Qi H et al. [7] and Luo YM et al. [6]. This may be due to different sequencing platforms and differences in the definition range of the borderline Z-score, resulting in a significant difference in the detection failure rate (borderline Z-score). After re-construction, 211 cases were indicated as the borderline Z-score, indicating a repeat sampling rate (borderline Z-score) of 0.25% (211/83,443), which is consistent with the aforementioned studies [6, 7]. After re-sampled, NIPT results indicated that 26 cases failed due to the borderline Z-score, accounting for 0.03% (26/83443). This demonstrates that library re-construction and/or re-sampled can greatly reduce the NIPT detection failure rate (borderline Z-score), decreasing it from 0.84 to 0.03%. FF has been shown to be significantly associated with the accuracy of NIPT and adverse pregnancy outcomes [15, 16]. This study analyzed the changes in FF across three sequential tests and observed a consistent increase in FF, accompanied by a continuous decrease in the proportion of borderline Z-score. A negative correlation was identified between these two trends (R² = 0.659). These findings suggest that repeated testing and advancing gestational age may contribute to an increase in FF, thereby improving the accuracy of NIPT and subsequently reducing the incidence of borderline Z-score. Ultimately, among the 700 pregnant women, 635 cases (90.71%, 635/700) received low-risk results, 34 cases (4.86%, 34/700) indicated abnormal NIPT results, 26 cases indicated NIPT detection failure, and 5 participants declined repeat sampling. Follow-up did not find any false negatives.These findings indicate that re-construction and/or re-drawing can screen out most normal pregnant women, effectively reducing unnecessary invasive prenatal diagnoses and alleviating the psychological burden on pregnant women.

In addition, this study found that the true-positive detection rate varies for different borderline Z-score types. The true-positive rate for the T13 borderline is the highest (1.3%), followed by the T21 borderline (1.2%) and the T18 borderline (0.4%), with the positive predictive value (PPV) also decreasing accordingly. Furthermore, the R² between mean FF and the borderline Z-score rate in the T18 borderline was 0.678, which is higher than that observed in the T21 (R² = 0.555) and T13 (R² = 0.437) borderline. This suggests that the T18 borderline is more susceptible to the influence of fetal fraction, resulting in the highest “false-positive rate” among the three trisomy screenings. When two or more borderline types coexist, no true positives were detected, mostly caused by data fluctuations. This may guide clinical genetic counseling for such populations. In addition, this study found that when NIPT results indicate abnormal results, pregnant women have good compliance with prenatal diagnosis, but they do not pay enough attention to the borderline Z-score, with 80.6% (25/31) of pregnant women not choosing prenatal diagnosis (see Table 3). Therefore, for pregnant women with indeterminate NIPT results due to Z-values falling within the borderline Z-score, it is recommended to actively refer them to qualified genetic counselors or obstetricians for comprehensive prenatal genetic counseling. Based on a thorough assessment of the patient’s individual situation—including serum biomarkers, previous ultrasound findings, and personal preferences—further prenatal diagnostic procedures should be recommended. These may include invasive diagnostic techniques (such as amniocentesis) and detailed systematic ultrasound screening. Through this approach, an individualized clinical decision-making plan can be formulated, thereby effectively reducing the risk of missed diagnosis of serious birth defects. Mitotic or meiotic non-disjunction errors can lead to the formation of mosaicism, resulting in the existence of two or more cell lines in the fetus and placenta [17]. Placental mosaicism is the main cause of false negatives and false positives in NIPT [18]. Among the five cases of fetal chromosomal abnormalities, two were mosaic types, including one case of translocation-type trisomy 21 mosaicism with a mosaic ratio of 40%, and another case of trisomy 13 mosaicism with a mosaic ratio of 43%. Both cases had two cell lines in the fetus, possibly due to meiotic errors forming a triploid zygote, and then through mitotic in subsequent mitosis, losing the extra chromosomes, ultimately leading to the coexistence of normal diploid and abnormal triploid [19]. Both cases initially showed borderline Z-score, and after library re-construction, the Z-score exceeded 3, showing that multiple NIPT results with Z-scores near 3 may indicate the possibility of fetal and placental mosaicism. The other three cases were other chromosomal abnormalities, possibly due to abnormal chromosome fragments affecting the calculation of Z-scores. In addition, one case of twins with one abnormality leading to borderline Z-scores was found. A 34-year-old pregnant woman transplanted two fresh embryos, and at 13 weeks of pregnancy, ultrasound indicated one twin with NT 5.3 mm, dotted nasal bone, and omphalocele. It is highly suspicious for trisomy 18, and the abnormal fragments released by the abnormal fetus neutralize with the normal fragments of the normal fetus, reducing the concentration of abnormal fragments and resulting in a borderline Z-score for trisomy 18 in NIPT.

In summary, among pregnant women with initial NIPT results indicating borderline Z-scores, both the screening positive rate and fetal abnormality rate are higher than those in the general population. Therefore, it is necessary to establish a borderline Z-score in NIPT to reduce the rate of missed screenings. In addition, library re-construction and/or re-drawing can significantly decrease the NIPT detection failure rate caused by the borderline Z-score, effectively identifying the majority of normal pregnancies and reducing unnecessary invasive prenatal diagnoses. For pregnant women who fail detection due to the borderline Z-score, active guidance should be provided to encourage further diagnostic testing to prevent missed screenings.

Conflict of interest

The authors declare no competing interests.

Author contributions

Xiaoli Pan and Lixin Weng are co-first authors of the article.Conceptualization, Haibo Li, Xiaoli Pan; methodology, Lixin Weng, Xiaoli Pan; formal analysis, Shuqing Pan, Shanshan Wu; investigation, Yun Pan, Shuqing Pan and Shanshan Wu; resources, Changshui Chen, Haibo Li; writing original draft preparation, Xiaoli Pan, Lixin Weng; writing-review and editing, Haibo Li; supervision, Changshui Chen; project administration, Changshui Chen, Haibo Li; funding acquisition, Changshui Chen, Haibo Li. All authors have read and agreed to the published version of the manuscript.

Funding

Ningbo key Laboratory of Genomic Medicine and Birth Defects Prevention, Key Technology Breakthrough Program of ‘Ningbo Sci-Tech Innovation YONGJIANG 2035, 2024Z221, Ningbo Top Medical and Health Research Program, 2022020405, and Ningbo key technology research and development project, 2023Z178.

Data availability

No datasets were generated or analyzed during the current study.