Fig 1.
Validation cohort of 717 unique samples evaluated in TSO500 analytical validation study.
A) Distribution of the 31 tumor types evaluated. Patients ranged in age from 5–99 years (406 female, 307 male, 4 unspecified). B) Summary of variant categories by tumor type.
Fig 2.
TSO500 workflow and supporting systems designed for a 70-sample run of routine FFPE tumor tissues.
A) CONNECT® supports order entry, specimen collection, sample accessioning, and results transmission through use of a Web Portal. B) Clarity LIMSTM manages the test specific lab workflow including Sample Preparation from Tissue QC through library preparation. C) High-throughput next-generation sequencing of 140 matched DNA/RNA libraries on NovaSeq 6000TM Sequencing System. D) The TSO500TM pipeline analyzes the genomic data to generate qualified variants. E) Variant annotation, review, treatment assignment, pathologist signout and clinical reporting in the GenomOncology Pathologist Workbench.
Fig 3.
TSO500 summary sequencing metrics for 250 FFPE samples using standard 70 paired sample sequencing run and 40ng DNA and 20ng RNA inputs.
A) DNA median coverage (UMI reads) across samples for all positions (blue) and Tier1 mutation positions with clinical significance (black line). B) DNA median coverage (UMI reads) across 9,232 TSO500 targeted genomic features. C) RNA mean and maximum coverage (blue bars) across 55 target genes.
Fig 4.
Concordance summary across all variant classes as positive percent agreement (PPA, green bar) and negative percent agreement (NPA, blue bar).
The analytical accuracy of TSO500 was evaluated using clinical FFPE samples obtained from patients with a variety of tumor types to assess agreement with an orthogonal validated NGS assay (vNGS). Data was aggregated at the variant level for substitutions, insertions and deletions, at the gene level for copy gain, copy loss, fusions and splice variants, and case level for MSI and TMB.
Fig 5.
Scatter Plot and correlation of TSO500 TMB score against (A) vNGS panel and (B) vWES.
Fig 6.
Precision summary across variable conditions including within (intra-) and between (inter-) runs, operators, days, instruments, reagent lots and barcodes.
A) TMB precision was assessed using %CV of the TMB score across replicates for samples bordering consensus the TMB-High (>10 Muts/Mb) cut off. B) Imprecision in variant detection assessed using average positive agreement (APA, orange) and average negative agreement (ANA, white) within replicates for all variant classes.
Table 1.
Analytical sensitivity (LoD VAF) for small variants (SNVs and indels).
Table 2.
Analytical sensitivity (LoD tumor purity) for copy gain, copy loss, MSI, TMB, fusions and splice variants.
Table 3.
Concordance summary for all variant classes and biomarkers.