Noninvasive detection of twin zygosity using genome-wide linkage disequilibrium information

Abstract

Dear Editor, in this study, we propose a novel linkage disequilibrium information-based noninvasive zygosity (LDNZ) assessment method in twin pregnancies. It combines fetus-specific allele frequency analysis with LD block to reduce the number of required single nucleotide polymorphism markers and experiment costs. LDNZ method offers a noninvasive, accurate, and cost-effective solution for zygosity assessment, addressing the need for precise obstetric care in twin pregnancies.

FIGURE 1

Schematic representation of linkage disequilibrium (LD)‐blocks on autosomal chromosomes. (A) The distribution of 1–22 autosomal chromosomes comprising 207 blocks, with red areas indicating block regions. Display linkage disequilibrium analysis plots for five single nucleotide polymorphism (SNP) markers within the chr11: 14160504–14202168 region, along with the probabilities of the two main haplotypes (hap1 and hap2). (B) The classification principle for SNP sites is based on maternal and corresponding paternal samples. Red squares represent hap1, while blue squares represent hap2.

FIGURE 2

Linkage disequilibrium information‐based noninvasive zygosity (LDNZ) for zygosity assessmnet and fetal fraction evaluation (A) Binomial distribution for the M042 twin. None of the Category 2 blocks met the conditions of p _bg ≤ .05 and p _Total ≤ .05. M042 was identified as the monozygotic twin. (B) Binomial distribution for M051 twin. 33 Category 2 blocks met the conditions of p _bg ≤ .05 and p _Total ≤ .05. M051 was identified as the dizygotic twin. (C) The distribution of total fetal fraction for 27 MZ twins and 28 DC twins. (D) The distribution of individual fetal fraction for 28 DC twins (ff1 < ff2). The difference between individual fetal fractions in M034 and M043 twins shows with multiples of 3.4‐fold (red highlight line) and 2.8‐fold (blue highlight line).

FIGURE 3

Histograms and receiver operating characteristic (ROC) analysis for linkage disequilibrium information‐based noninvasive zygosity (LDNZ) method and classical single nucleotide polymorphism (SNP)‐based method (A) Histogram showing the distribution of the number of informative blocks identified by the LDNZ analysis. (B) ROC curve analysis for the number of informative blocks identified by the LDNZ method. At the optimal model prediction performance, the best cutoff value is 2, with specificity, sensitivity, and area under the curve (AUC) equal to 1. (C) Histogram showing the distribution of the number of informative loci identified by the SNP analysis. (D) ROC curve analysis for the number of informative SNPs identified by the SNP method. At the optimal model prediction performance, the best cutoff value is 28, with a specificity of 1.00, a sensitivity of 0.93, and an AUC of 0.989.