Skip to main content

Table 5 RMSE, precision, and recall obtained by Flimma and the meta-analysis tools on TCGA-BRCA datasets split by tissue source sites

From: Flimma: a federated and privacy-aware tool for differential gene expression analysis

The number of cohorts 3 5 7 10 14
RMSE
 Flimma 0.0008 0.0007 0.0008 0.0017 0.0012
 Fisher 0.94 1.82 2.53 3.86 5.37
 Stouffer 1.47 2.21 2.87 4.26 5.68
 REM 2.73 3.68 4.75 7.21 8.50
 RankProd 5.16 8.19 11.32 18.92 23.50
Precision
 Flimma 1.00 1.00 1.00 1.00 1.00
 Fisher 0.85 0.88 0.90 0.93 0.95
 Stouffer 0.85 0.88 0.91 0.93 0.95
 REM 0.93 0.94 0.95 0.97 0.97
 RankProd 0.92 0.87 0.90 0.93 0.95
Recall
 Flimma 1.00 1.00 1.00 1.00 1.00
 Fisher 0.92 0.95 0.95 0.96 0.97
 Stouffer 0.89 0.93 0.94 0.96 0.97
 REM 0.93 0.96 0.97 0.98 0.98
 RankProd 0.87 0.96 0.96 0.96 0.97
  1. Values corresponding to the best performance over all methods are italicized