Systematic Evaluation of Sanger Validation of Next-Generation Sequencing Variants

Abstract

Background: Next-generation sequencing (NGS) data are used for both clinical care and clinical research. DNA sequence variants identified using NGS are often returned to patients/participants as part of clinical or research protocols. The current standard of care is to validate NGS variants using Sanger sequencing, which is costly and time-consuming.

Methods: We performed a large-scale, systematic evaluation of Sanger-based validation of NGS variants using data from the ClinSeq® project. We first used NGS data from 19 genes in 5 participants, comparing them to high-throughput Sanger sequencing results on the same samples, and found no discrepancies among 234 NGS variants. We then compared NGS variants in 5 genes from 684 participants against data from Sanger sequencing.

Results: Of over 5800 NGS-derived variants, 19 were not validated by Sanger data. Using newly designed sequencing primers, Sanger sequencing confirmed 17 of the NGS variants, and the remaining 2 variants had low quality scores from exome sequencing. Overall, we measured a validation rate of 99.965% for NGS variants using Sanger sequencing, which was higher than many existing medical tests that do not necessitate orthogonal validation.

Conclusions: A single round of Sanger sequencing is more likely to incorrectly refute a true-positive variant from NGS than to correctly identify a false-positive variant from NGS. Validation of NGS-derived variants using Sanger sequencing has limited utility, and best practice standards should not include routine orthogonal Sanger validation of NGS variants.

Figure 1. Box-And-Whisker Plots of Gene Statistics

Box-and-whisker plots show the distribution of genes in each candidate set used in this analysis across GC content (Figure 1A), CDS length (Figure 1B), and exon count (Figure 1C). Data on these genes was collected using UCSC Genome Browser (28) or NCBI’s Entrez (29). In the case of multiple transcripts, the transcript encoding the longest protein isoform was used.

Figure 2. Jaccard Index with Increasing MPG Cutoff

The Jaccard sameness index was used to evaluate the agreement of variants discovered using NGS versus Sanger sequencing, correlated with increasing MPG thresholds, then plotted using R. Data from both sequencing methods were in complete agreement at an MPG higher than 10, resulting in a score of 1.000 at all points thereafter.