Skip to main content

Table 1 Tools comparison on simulated data for k = 1

From: KAOS: a new automated computational method for the identification of overexpressed genes

 

k = 1, T = 10

k = 1, T = 20

k = 1, T = 50

 

Precision

Recall

F-measure

Precision

Recall

F-measure

Precision

Recall

F-measure

KAOS

0.348

0.348

0.348

0.267

0.267

0.267

0.162

0.162

0.162

Zodet

0.232

0.232

0.232

0.243

0.243

0.243

0.220

0.220

0.220

GTI

0.182

0.182

0.182

0.175

0.175

0.175

0.146

0.146

0.146

Khotary et al.

-

0.038

-

0.114

0.032

0.050

0.121

0.012

0.022

  1. The comparison of Kaos performances is based on 50 simulations on a synthetic dataset made of 1000 genes expression values for 30 cases and 30 cancer test samples. The expression values were drawn from a normal distribuion with mean 7 and standard deviation 1, where k samples which have been marked as outliers’ samples (see Methods section for further details) and T is the top T number of outlier genes found. The table shows average Precision, Recall and F-Measure for k =1 and t ranging from 10 to 50
  2. In bold are reported the values obtained by the best performing tool in the different conditions
Back to article page

BMC Bioinformatics

ISSN: 1471-2105

Contact us