|
Performance of the CNVfinder algorithm using validated data before wave correction |
||||||
| Calls |
||||||
|
|
||||||
| Status |
Number of regions |
A |
B |
C |
D |
E |
|
|
||||||
| Unvalidated |
25 |
4 |
4 |
3 |
0 |
0 |
| Validated |
112 |
68 |
70 |
65 |
47 |
35 |
| Total |
137 |
72 |
74 |
68 |
47 |
35 |
| False discovery rate |
6 |
5 |
4 |
0 |
0 |
|
| Sensitivity (confidence interval) |
61 (51, 70) |
62 (53, 72) |
58 (48, 67) |
42 (33, 52) |
31 (23, 41) |
|
| Specificity (confidence interval) |
84 (66, 95) |
84 (66, 95) |
88 (71, 97) |
100 (90, 100) |
100 (90, 100) |
|
|
The number of regions called in five replicate experiments (A to E, ranked by standard deviation) using CNVfinder before wave correction. The false discovery rate, sensitivity and specificity are quoted as percentages and the confidence intervals quoted are the 95% intervals. | ||||||
Marioni et al. Genome Biology 2007 8:R228 doi:10.1186/gb-2007-8-10-r228 |
||||||