An Optimized Method for Accurate Fetal Sex Prediction and Sex Chromosome Aneuploidy Detection in Non-Invasive Prenatal Testing

Abstract

Massively parallel sequencing (MPS) combined with bioinformatic analysis has been widely applied to detect fetal chromosomal aneuploidies such as trisomy 21, 18, 13 and sex chromosome aneuploidies (SCAs) by sequencing cell-free fetal DNA (cffDNA) from maternal plasma, so-called non-invasive prenatal testing (NIPT). However, many technical challenges, such as dependency on correct fetal sex prediction, large variations of chromosome Y measurement and high sensitivity to random reads mapping, may result in higher false negative rate (FNR) and false positive rate (FPR) in fetal sex prediction as well as in SCAs detection. Here, we developed an optimized method to improve the accuracy of the current method by filtering out randomly mapped reads in six specific regions of the Y chromosome. The method reduces the FNR and FPR of fetal sex prediction from nearly 1% to 0.01% and 0.06%, respectively and works robustly under conditions of low fetal DNA concentration (1%) in testing and simulation of 92 samples. The optimized method was further confirmed by large scale testing (1590 samples), suggesting that it is reliable and robust enough for clinical testing.

Fig 1. Discovery of six special regions in Y chromosome.

A and B show the six highly covered regions (highlighted by red line) in Y chromosome in female and male group. They were located at the same positions of Y chromosome in two groups. C) There are detection counts of six regions in Y chromosome of 92 samples. D) The comparison of Y chromosome reads percentage in the female and male fetus groups in before and after filtering out the reads in six regions of Y chromosome.

Fig 2. Testing of the optimized method in large-scale sample.

A) The DNA percentages of Y chromosome calculated by optimized method and original method. Here, the fetal DNA concentration is assumed as 1%. The percentage of Y chromosome calculated by the optimized method was marked as green point. The blue point illustrates the percentage of Y chromosome calculated by original method. B and C) the percentage of Y chromosome calculated by optimized method and original method in 1590 samples. Here, black and red dots represent the percentage of Y chromosome in female fetuses and male fetus, respectively. Error prediction results were marked by blue and green dots in Fig C in which two male fetuses were predicted to be female and marked by blue; a female fetus predicted to be male and marked by green.