NextGen Sequencing Analytic Validity
As of March 2016, 6.36 Mb of sequence (83 genes, 1557 exons) generated in our lab was compared between Sanger and NextGen methodologies. We detected no differences between the two methods. The comparison involved 6400 total sequence variants (differences from the reference sequences). Of these, 6144 were nucleotide substitutions and 256 were insertions or deletions. About 65% of the variants were heterozygous and 35% homozygous. The insertions and deletions ranged in length from 1 to over 100 nucleotides.
In silico validation of insertions and deletions in 20 replicates of 5 genes was also performed. The validation included insertions and deletions of lengths between 1 and 100 nucleotides. Insertions tested in silico: 2200 between 1 and 5 nucleotides, 625 between 6 and 10 nucleotides, 29 between 11 and 20 nucleotides, 25 between 21 and 49 nucleotides, and 23 at or greater than 50 nucleotides, with the largest at 98 nucleotides. All insertions were detected. Deletions tested in silico: 1813 between 1 and 5 nucleotides, 97 between 6 and 10 nucleotides, 32 between 11 and 20 nucleotides, 20 between 21 and 49 nucleotides, and 39 at or greater than 50 nucleotides, with the largest at 96 nucleotides. All deletions less than 50 nucleotides in length were detected, 13 greater than 50 nucleotides in length were missed. Our standard NextGen sequence variant calling algorithms are generally not capable of detecting insertions (duplications) or heterozygous deletions greater than 100 nucleotides. Large homozygous deletions appear to be detectable.
Sanger Sequencing Analytical Validity
As of February 2018, we compared 26.8 Mb of Sanger DNA sequence generated at PreventionGenetics to NextGen sequence generated in other labs. We detected only 4 errors in our Sanger sequences, and these were all due to allele dropout during PCR. For Proficiency Testing, both external and internal, in the 14 years of our lab operation we have Sanger sequenced roughly 14,300 PCR amplicons. Only one error has been identified, and this was an error in analysis of sequence data.
Our Sanger sequencing is capable of detecting virtually all nucleotide substitutions within the PCR amplicons. Similarly, we detect essentially all heterozygous or homozygous deletions within the amplicons. Homozygous deletions which overlap one or more PCR primer annealing sites are detectable as PCR failure. Heterozygous deletions which overlap one or more PCR primer annealing sites are usually not detected (see Analytical Limitations). All heterozygous insertions within the amplicons up to about 100 nucleotides in length appear to be detectable. Larger heterozygous insertions may not be detected. All homozygous insertions within the amplicons up to about 300 nucleotides in length appear to be detectable. Larger homozygous insertions may masquerade as homozygous deletions (PCR failure).