Efficient Methods to Generate NIPT Male Samples with Aberrations on Sex Chromosomes for Benchmarking NIPT Algorithms

Noninvasive prenatal test (NIPT) has become a routine screening method to detect autosomal chromosomal aberrations. Predicting sex chromosome aneuploidies (SCAs) is more difficult due to the complicated genomic characteristics of sex chromosomes. Benchmarking the sensitivity of NIPT algorithms is challenging due to the lack of positive samples. In this paper, we introduce simple and efficient computational methods to create aberrations on sex chromosomes for NIPT male samples (i.e., creating male samples with XXY or XYY aberrations) from negative NIPT samples. We applied the methods to create 513 positive samples with XXY aberrations and 200 positive samples with XYY aberrations to evaluate the sensitivity of three NIPT algorithms WisecondorX, its improvement VINIPT, and CNVkit. Experiements on both negative and positive samples showed that the algorithms have high sensitivity in detecting XXY and XYY aberrations (100% for WisecondorX and VINIPT; and 99.2% for CNVkit). The VINIPT algorithm has an overall specificity of 99.8% and better than WisecondorX (∼96.8%) and CNVkit (∼99%). The results indicate that VINIPT can play as a powerful tool for screening aberrations on sex chromosomes of male samples.