Fig. 5: Assessing copy number variations.
From: Clinical and analytical validation of a combined RNA and DNA exome assay across a large tumor cohort

a Schema for CNV calling using BAF and tumor/normal depth ratio. b Whole exome CNV calling examples for reference cell lines with varying ploidy (2n, 3n, 4n). CNV interpretation depends on tumor ploidy. c Sensitivity of CNV calling for COLO829 cell line (n = 5692) and its dependence on purity. d Comparison of COLO829 exome from FF and FFPE tissues. e BAF and depth ratio correlation between FF and FFPE samples. Shading shows the 95% confidence interval. f Arm- and gene-level CNV and g LOH/cnLOH calling performance across different purities for five reference cell lines (error bars show s.e.m.). h Heatmaps of exome-wide CNV comparison across 8 clinical samples sequenced in different labs; top - high purity, bottom - low purity. Gray - neutral, blue - shallow deletion, purple - deletion, pink - gain, red - amplification. i Concordance of CNV calling with orthogonal sequencing in high purity (>30%) clinical samples (n = 39). Box plots show the interquartile range.