Table 2 |
||
|
Number of MAQC data sets for which each estimator ranked in the top four. |
||
|
algorithm |
cross validation posterior |
predictive expected loss |
|
|
||
|
Wilcoxon stat. with emp. null |
0 |
0 |
|
|
||
|
t stat. with emp. null |
0 |
0 |
|
|
||
|
Wilcoxon stat. with locfdr |
0 |
0 |
|
|
||
|
t stat. with locfdr |
1 |
3 |
|
|
||
|
moderated t stat. with locfdr |
3 |
3 |
|
|
||
|
moderated t stat. with limma |
2 |
0 |
|
|
||
|
fold change shrinkage |
0 |
0 |
|
|
||
|
Bayes factor |
3 |
0 |
|
|
||
|
BIC |
2 |
3 |
|
|
||
|
AICc |
1 |
3 |
|
|
||
|
fold change > 1.41 |
0 |
0 |
|
|
||
|
fold change > 1 |
0 |
0 |
|
|
||
|
fold change > 2 |
0 |
0 |
|
|
||
|
all nulls true |
0 |
0 |
|
|
||
|
all nulls false |
0 |
0 |
|
|
||
|
Yanofsky and Bickel BMC Bioinformatics 2010 11:63 doi:10.1186/1471-2105-11-63 |
||
