Skip to main content

Table 3 Results obtained with the newly proposed gene sets on the selection data sets

From: Novel gene sets improve set-level classification of prokaryotic gene expression data

Gene-set type

Mean accuracy [%]

Sum of ranks

Operon (OPR)

81.50

196.00

Transcriptional unit (TU)

82.17

198.50

Continuous subsequence (COPR)

83.00

205.50

Transcription factor (TF)

78.69

179.00

Regulon (REG)

82.67

211.50

Strict regulon (SREG)

82.50

209.50

  1. Columns contain the mean accuracies and sum-of-ranks indicators over the datasets, higher rank indicates better performance. Here, the best ranked gene-set types from the two categories (operon-based, transcription-factor based) are COPR and REG, respectively