 |
 |
 |
 |
 |
 |
 |
 |
 |
 |
 |
 |
 |
 |
 |
 |
 |
 |
|
calculate the t-statistics for each gene in the real data
|
|
|
|
generate bootstrapped data sets by permuting the
|
|
|
columns (conditions)--preserves dependence
|
|
|
between
genes
|
|
|
The highest-ranking gene is up- (down-) regulated, if
|
|
for at least 95% (99%) of the bootstrapped
|
|
|
datasets, all genes have a smaller t-value. This
|
|
|
gene is then removed from the set and the
|
|
|
definition is applied recursively until the j-th
|
|
|
ranked gene is no longer deemed to be
|
|
|
differentially regulated.
|
|