Quality control (QC) thresholds

QC Test

Detection threshold

SNV/Indel

A limit of detection (LOD) is calculated and displayed for each variant call. LOD is determined by the level of molecular amplicon coverage. If no variant call is detected, the LOD range is displayed across entire amplicon.

Molecular coverage must be at least 2 with a minimum detection cutoff frequency of 0.035% and 0.05% for lung and breast panels, respectively.

CNV

The MAPD metric is a measure of read coverage noise detected across all amplicons in a panel. Higher MAPD typically translates to lower coverage uniformity. Lower coverage uniformity can result in missed or erroneous CNV calls. MAPD score is viewable in downloadable VCF file or review of the Analysis Results of a single sample extended analysis.

To make a CNV call the following criteria must be met:

  • MAPD <0.4

  • P-value <10-5

  • CNV Ratio for a copy number gain must be >1.15

  • CNV Ratio for a copy number loss must be <0.85

Note: The CNV Ratio call thresholds were derived empirically using plasma samples from healthy donors with normal CNV status.

Fusions/Exon Skipping[1]

  • Fusions—Panel includes 2 process control target genes, TBP and HMBS. At least 1 control must have a molecular count of >2 to pass QC.

  • MET Exon Skipping—Panel includes 2 MET Wild Type control amplicons (gene name has WT at the end). At least 1 of these controls must have a molecular count >2 to pass QC.

Fusion and Exon Skipping amplicons must have >2 molecular counts to be reported.

1 These variant types are included in the Oncomine™ Lung cfNA Assay, derived from RNA reverse‑transcribed into cDNA during library preparation.