Table 2. Results of simulations varying coverage and sample numbers.
Simulations of DECoN exon CNV calling in the evaluation set, using different coverage and sample numbers, were performed. Average sensitivity and specificity values are shown, with the range in parentheses. Sensitivity and specificity achievable by an Illumina sequencing run generating 15M read pairs (typical of a MiSeq run) and 300M read pairs (typical of a HiSeq2500 Rapid run) are shown in green and blue, respectively.
| Number of samples per sequencing pool | ||||||
|---|---|---|---|---|---|---|
|
Coverage per
sample in million (M) read pairs |
6 | 12 | 24 | 48 | 96 | |
| 1.25 | Sensitivity | 87% (0–100) | 92% (40–100) | 92% (60–100) | 91% (81–100) | 96% (94–100) |
| Specificity | 98% (56–100) | 99% (89–100) | 99% (96–100) | 99% (98–100) | 99% (99–100) | |
| 2.5 | Sensitivity | 95% (0–100) | 98% (67–100) | 99% (83–100) | 100% (94–100) | 100% (100–100) |
| Specificity | 97% (50–100) | 99% (89–100) | 99% (95–100) | 99% (98–100) | 99% (99–99) | |
| 3.125 | Sensitivity | 99% (67–100) | 99% (75–100) | 100% (100–100) | 100% (100–100) | 100% (100–100) |
| Specificity | 97% (50–100) | 99% (94–100) | 100% (97–100) | 99% (99–100) | 99% (99–99) | |
| 5 | Sensitivity | 96% (0–100) | 99% (75–100) | 100% (100–100) | 100% (100–100) | 100% (100–100) |
| Specificity | 97% (50–100) | 99% (84–100) | 99% (93–100) | 99% (96–100) | 99% (98–99) | |
| 7.5 | Sensitivity | 96% (0–100) | 99% (75–100) | 100% (89–100) | 100% (94–100) | 100% (100–100) |
| Specificity | 97% (56–100) | 99% (84–100) | 99% (92–100) | 99% (96–100) | 99% (98–99) | |