Evaluation of false positive and false negative errors in targeted next generation sequencing.

Area of Science:

• Genomics
• Molecular Diagnostics
• Bioinformatics

Background:

• Next-generation sequencing (NGS) is vital for disease diagnostics.
• Sequencing and analysis errors hinder clinical use, particularly in cancer diagnostics for low-level somatic variants due to tumor heterogeneity.

Purpose of the Study:

• To assess the performance variability of targeted next-generation sequencing (T-NGS) across certified providers and analysis pipelines.
• To identify factors contributing to sensitivity differences and false positive (FP) error rates in T-NGS.

Main Methods:

• Utilized reference-standard DNA mixtures with varying ratios of homozygote hydatidiform mole and heterozygote blood DNA.
• Analyzed samples using T-NGS by certified providers and compared different analysis pipelines (DRAGEN, BWA+GATK Mutect2, in-house).
• Validated discordant variant calls using single base extension assays.

Main Results:

• Significant variability observed in analytical sensitivity (up to 13.9-fold) and FP error rates (up to 615-fold) among providers and pipelines.
• Identified recurrent FP-prone alleles as a major contributor to total FP errors in specific pipelines.
• Conventional pipelines (BWA+GATK Mutect2) showed increased FP errors compared to DRAGEN, despite similar sensitivity.

Conclusions:

• Certified T-NGS providers exhibit substantial variability in sensitivity and FP error rates.
• Conventional analysis pipelines increase FP errors and accumulate recurrent FP-prone alleles.
• Standardized pipelines and rigorous quality control are essential for reliable clinical T-NGS diagnostics.