Quality threshold evaluation of Sanger confirmation for results of whole exome sequencing in clinically diagnostic setting

Abstract

BackgroundWith the ability to simultaneously sequence more than 5,000 disease-associated genes, next-generation sequencing (NGS) has replaced Sanger sequencing as the preferred method in the diagnostic field at the laboratory level. However, Sanger sequencing has been used routinely to confirm identified variants prior to reporting results. This validation process causes a turnaround time delay and cost increase. Thus, this study aimed to set a quality threshold that does not require Sanger confirmation by analyzing the characteristics of identified variants from whole exome sequencing (WES). MethodsOur study analyzed data on a total of 694 disease-causing variants from 578 WES samples that had been diagnosed with suspected genetic disease. These samples were sequenced by Novaseq6000 and Exome Research Panel v2. All 694 variants (513 single-nucleotide variants (SNVs) and 181 indels) were validated by Sanger sequencing. ResultsA total of 693 variants included 512 SNVs and 181 indels from 578 patients and 367 genes. Five hundred seven heterozygous SNVs with at > 250 quality score and > 0.3 allele fraction were 100% confirmed by Sanger sequencing. Five heterozygous variants and one homozygous variant were not confirmed by Sanger sequencing, which showed 98.8% accuracy. There were 146 heterozygous variants and 35 homozygous variants among 181 indels, of which 11 heterozygous variants were not confirmed by Sanger sequencing (93.9% accuracy). Five non-confirmed variants with high quality were not identified on the ram .bam file. ConclusionOur results indicate that Sanger confirmation is not necessary for exome-derived SNVs with > 250 quality score and 0.3 > allele fraction set to an appropriate quality threshold. Indels or SNVs that do not meet the quality threshold should be reviewed by raw .bam file and Sanger confirmation should be performed to ensure accurate reporting.