Table 2.
PG1 | 50% | 70% | 90% |
---|---|---|---|
(a) TMM |
|
|
|
PDEG = 5% |
0.13 |
0.18 |
0.22 |
PDEG = 15% |
0.16 |
0.19 |
0.15 |
PDEG = 25% |
0.17 |
0.14 |
0.15 |
(b) DEGES/TbT |
|
|
|
PDEG = 5% |
1492.92 |
1423.55 |
1573.23 |
PDEG = 15% |
1556.93 |
1510.17 |
1408.46 |
PDEG = 25% |
1527.08 |
1483.09 |
1543.80 |
(c) DEGES/edgeR |
|
|
|
PDEG = 5% |
3.04 |
3.08 |
3.24 |
PDEG = 15% |
3.08 |
3.05 |
2.89 |
PDEG = 25% |
2.99 |
2.92 |
3.05 |
(d) iDEGES/edgeR |
|
|
|
PDEG = 5% |
8.80 |
8.95 |
9.53 |
PDEG = 15% |
8.94 |
8.91 |
8.39 |
PDEG = 25% |
8.75 |
8.39 |
8.92 |
(e) iDEGES/TDT |
|
|
|
PDEG = 5% |
17.92 |
17.54 |
18.24 |
PDEG = 15% |
19.85 |
19.45 |
18.74 |
PDEG = 25% |
21.17 |
20.74 |
21.16 |
(f) iDEGES/DESeq |
|
|
|
PDEG = 5% |
17.88 |
17.54 |
18.34 |
PDEG = 15% |
19.96 |
19.43 |
18.72 |
PDEG = 25% | 21.17 | 20.75 | 21.27 |
Average computation times (in seconds) of 100 trials for the six normalization methods in Table 1 are shown. The results of DEGES/TbT were obtained by using a suggested parameter setting for performing bootstrap resampling (i.e., samplesize = 10000).