Fetal Fraction Signatures: A Quality Control Tool to Detect Potentially Confounding Situations in NonInvasive...

Noninvasive prenatal diagnosis relies on the analysis of the small amount of cell-free fetal DNA circulating in maternal plasma. Widely used to screen for chromosomal anomalies, the technique can also be applied to Mendelian diseases (NIPD-M). However, asserting the presence of a maternal variant in the fetus is challenging and requires the determination of the maternal haplotypes transmitted to the fetus. This is achieved via relative haplotype dosage (RHDO), a method that relies on the analysis of the allelic balance of multiple single-nucleotide polymorphisms around the pathogenic variant. An inherent risk of this method is that twin pregnancies and chromosomal anomalies like trisomy, monosomy or uniparental disomy unavoidably alter allelic balance, potentially leading to incorrect diagnosis. Here, we introduce an analytical method, fetal fraction signatures, that can detect such confounding situations in NIPD-M data, without the need for additional wet-lab work. We show that our method can reliably detect anomalies down to a low level of mosaicism and a low fetal fraction, thereby providing an essential quality-control tool for clinical testing by NIPD-M.