Fig 1.
Overview of the InVisionFirst workflow.
Fig 2.
Sensitivity for SNVs (A, C) and indels (B, D).
A and B show the sensitivity as a function of the allele fraction of the reference mutations. Each line represents a different operator/Lot combination. C and D show the full set of calls for all combinations of dilution/variant (vertical) and repeat/operator/lot (horizontal). Blue rectangles represent mutations that were detected and grey represents those missed. Panel E shows for SNVs the estimated allele fraction compared between InVisionFirst (blue box-plots) and the reference as estimated by Horizon using ddPCR (red crosses).
Fig 3.
Blue rectangles represent fusions that were detected and grey represents those missed. (A) Dilution of Horizon reference material containing 2 fusions (ALK and ROS1) across dilution levels (vertical) and operator/lot (horizontal), (B) Set 1 of contrived material based on published DNA breakpoints (AF 1% and 0.5%, 2 operators), (C) Set 2 of contrived material based on published DNA breakpoints (AF 1% and 0.5%, 2 operators, 2 reagent lots) and (D) Contrived material based on randomly generated fusion breakpoints. Different operators performed different parts of this fusion study.
Table 1.
Amplification sensitivity analysis for FGFR1, EGFR, ERBB2 and MET.
Fig 4.
Comparison between InVisionFirst assay and orthogonal dPCR generated by an independent laboratory.
Four common cancer mutations were tested by dPCR in 20 samples selected to have one of these four mutations based on the InVisionFirst assay. The allele fraction of the mutation not detected by the orthogonal method is estimated by the InVisionFirst assay at 0.3%.
Fig 5.
Stability of AF over time using Streck blood tubes spiked with Horizon reference material.