Cost-effective and accurate method of measuring fetal fraction using SNP imputation

Abstract

Motivation With the discovery of cell-free fetal DNA in maternal blood, the demand for non-invasive prenatal testing (NIPT) has been increasing. To obtain reliable NIPT results, it is important to accurately estimate the fetal fraction. In this study, we propose an accurate and cost-effective method for measuring fetal fractions using single-nucleotide polymorphisms (SNPs). Results A total of 84 samples were sequenced via semiconductor sequencing using a 0.3× sequencing coverage. SNPs were genotyped to estimate the fetal fraction. Approximately 900 000 SNPs were genotyped, and 250 000 of these SNPs matched the semiconductor sequencing results. We performed SNP imputation (1000Genome phase3 and HRC v1.1 reference panel) to increase the number of SNPs. The correlation coefficients (R 2) of the fetal fraction estimated using the ratio of non-maternal alleles when coverage was reduced to 0.01 following SNP imputation were 0.93 (HRC v1.1 reference panel) and 0.90 (1000GP3 reference panel). An R 2 of 0.72 was found at 0.01× sequencing coverage with no imputation performed. We developed an accurate method to measure fetal fraction using SNP imputation, showing cost-effectiveness by using different commercially available SNP chips and lowering the coverage. We also showed that semiconductor sequencing, which is an inexpensive option, was useful for measuring fetal fraction. Availability and implementation python source code and guidelines can be found at https://github.com/KMJ403/fetalfraction-SNPimpute. Contact kangskim@ajou.ac.kr or sunshinkim3@gmail.com Supplementary informationSupplementary dataare available at Bioinformatics online.