A new noninvasive prenatal paternity test based on double nucleotide polymorphisms abates sequencing uncertainties and improves sensitivity

Noninvasive prenatal paternity testing (NIPAT) based on fetal cell-free DNA (fetal cfDNA) has gained increasing relevance in forensic genetics. However, its applicability and reliability are often limited by the low fetal DNA fraction and by high-throughput sequencing errors that can be misinterpreted as inherited paternal variants. Here, we present NEWPAT, a novel NIPAT approach based on Double Nucleotide Polymorphisms (DNPs), i.e. biallelic variants composed of two adjacent nucleotides arising from a single mutational event. These markers offer a key advantage compared to single nucleotide polymorphisms (SNPs): the likelihood of both positions being affected by independent sequencing errors is extremely low, thus improving the specificity of fetal DNA detection. We develop a targeted amplicon-based panel including 979 markers (845 autosomal and 80 X Chromosome DNPs, plus 54 Y SNPs), and test at depths ranging from 1,000× to 10,000× on 42 individuals, including genomic DNA from 4 families, 2 artificial maternal-child mixtures simulating cfDNA and actual cfDNA from 12 pregnant women plus the putative biological fathers. We observe a DNP sequencing error rate of 8.9×10 -5 , i.e. ~12 times lower than the observed single base error rate. By exploiting the full range of sequencing depths at loci supporting the fetal genotype, we obtain Combined Paternity Index curves that consistently lead to the correct paternity attribution or exclusion in all true or false mother-father pairs. Notably, NEWPAT proves effective even at fetal cfDNA levels as low as 0.2%, when existing SNP-based protocols lead to inconclusive results, offering a highly sensitive and accurate NIPAT strategy.