|
Performance of the five methods in simulation |
||||||
| Method of Analysis |
||||||
|
|
||||||
| Percentage arrays damaged |
Single Channel |
Unaffected Data |
Inclusive |
Combined (kest) |
Combined (kfix = 0.5) |
Combined (kfix = 0.25) |
|
|
||||||
| 20 |
85,780 |
88,197 |
86,899 |
88,393 |
87,763 |
88,017 |
| 40 |
85,776 |
86,592 |
84,557 |
88,296 |
87,317 |
87,704 |
| 60 |
85,799 |
83,958 |
82,650 |
88,054 |
86,765 |
87,402 |
| 80 |
85,893 |
78,134 |
80,411 |
87,252 |
86,016 |
86,848 |
| All Simulations |
343,248 |
336,881 |
334,517 |
351,995 |
347,861 |
349,971 |
|
Presented are the numbers of differentially expressed genes that would be identified by taking a list of the top 4,000 genes in each case. Recall that there are indeed 4,000 differentially expressed genes, so the numbers are out of 100,000 (25 times 4,000) for the sub experiments and out of 400,000 (100 times 4,000) overall. | ||||||
Lynch et al. BMC Bioinformatics 2007 8:26 doi:10.1186/1471-2105-8-26 |
||||||