Enhancing routine noninvasive prenatal testing with cell-free DNA end characteristics

Abstract

Objective: Noninvasive prenatal testing (NIPT) based on cell-free DNA (cfDNA) analysis is widely used for detecting fetal aneuploidies globally, such as Trisomy 21 (T21) testing. Despite its high sensitivity, current NIPT methods have a non-negligible rate of false-negative results, primarily due to low fractions of fetal-derived cfDNA in maternal plasma, posing challenges for affected families and public health.

Methods and Analysis: We propose a computational approach to enrich fetal-derived cfDNA by leveraging end characteristics that does not require any modifications to existing experimental protocols. We have evaluated this method using three independent datasets, including over 2,200 samples from diverse ethnic backgrounds and experimental platforms.

Results: Here we show that through end selection, we significantly increase Z-scores in all T21 samples from 3 independent datasets, which shows potential in reducing false-negative results while not introducing any false positives in the euploid samples. Our method is compatible with current routine NIPT workflows that generate low-depth, short, and single-end whole genome sequencing data, therefore allowing for seamless integration with minimal additional cost.

Conclusion: Our method offers translational potential for enhancing routine NIPT by reducing the false negatives, addressing a critical limitation in current clinical practice.