Fig 1.
A. General view. Each dot corresponds to an amplicon. The amplicons are distributed on the x-axis according to their genomic position. The y-axis corresponds to the normalized values. Grey dots indicate a “normal” value, whereas red or orange dots indicate duplicated and deleted amplicons, respectively. The names of the gene and chromosome number are located at the bottom of the figure. The green curve shows the Loess regression. The thick green ribbon is a noise heatmap in which green indicates a stable amplicon in all samples (see S1 File). The red rectangle highlights a CNV region. B. Zoom on the duplicated region covered by 10 amplicons (PMP22).
Fig 2.
Comparison of single-stage and two-stage ratio results.
A. Without the two-stage ratio, a disturbed region showed a false-positive duplication on chromosome 13 covered by three amplicons. B. The two-stage ratio improved the stability of the region and the false duplication was no longer detected. C. Without the two-stage ratio, six amplicons (grey dots in Chr13 area) were not detected as deleted throughout chromosome 13 (39 amplicons) and three separated CNVs were detected. D. With the two-stage ratio, only three false-negative amplicons (grey dots in chr13 area) were present among the 39 amplicons of chromosome 13 and only one amplicon split the total deletion of the chromosome (partial screenshots from CovCopCan).
Fig 3.
Example of CNV merging on a chromosome X duplication.
A. Entire duplication of chromosome X. CovCopCan detects six CNV areas without the merging CNV algorithm. B. By using the merging CNV algorithm, the duplication detected includes all of chromosome X, although some amplicons appear as neutral (grey dots). C. The exported CNV in the VCF format contains only one line corresponding to the duplication of chromosome X (partial screenshots from CovCopCan).
Fig 4.
Visualization of chromosome 13 in seven samples.
Each dot corresponds to an amplicon. Orange and red rectangles correspond to deletions and duplications, respectively. The green curve shows the Loess regression. Patients 1 to 4 share a q arm deletion. Samples 5 to 7 do not present this deletion. Without defining samples 5 to 7 as controls, only one deletion was correctly detected in patient 4. A partial deletion was detected in patient 1. False-positive deletions were detected in two of the three controls. By defining samples 5 to 7 as controls, two deletions were correctly detected in patients 1 and 4. Two partial deletions were found in both patients 3 and 4. No duplication was found in the controls (partial screenshots from CovCopCan).
Table 1.
Details of the 22 positive-control CNVs used for germline analysis, with chromosomal locations of the CNVs.
a: Number of amplicons covering the CNVs. b: Number of amplicons correctly detected as duplicated or deleted by CovCovCan.
Table 2.
Comparison of the performance of CovCopCan and other CNV callers for 22 positive-control CNVs from 22 samples.
Fig 5.
Gene deletion simulation (gene visualized in red), with various proportions of cells containing this deletion.
The cumulative summary chart (blue shading) first detected the deletion with 15 to 20% of the cells containing the deletion (partial screenshots from CovCopCan).
Fig 6.
Detection of the entire ATM gene deletion in patients DNA, in which the percentage of cancer cells was estimated based on 200 FISH metaphases per patient.
The Cumulative summary detected the deletion starting from 19.5% estimated cancer cells (partial screenshots from CovCopCan).
Table 3.
Detection of a CNV according to the proportion of cancer cells.
“No” indicates no detection of the CNV, whereas “Yes” indicates correct detection of the CNV.
Fig 7.
Deletion of the arm of chromosome 13 detected by CovCopCan using the Cumulative Summary Chart.
The deletion is highlighted in the blue area.