Skip to main content
. 2015 Nov 20;9:84. doi: 10.1186/s12918-015-0233-4

Table 2.

Comparison of MEN, BBSR and PriorPC in terms of AUROC

DREAM4 E. coli B.subtilis Using TS
MEN_low 0.908 0.768 0.828 Yes
MEN_high 0.912 0.776 0.842 Yes
BBSR_low 0.872 0.675 0.791 Yes
BBSR_high 0.86 0.719 0.793 Yes
PriorPC (α=1) 0.887 0.753 0.835 No
PriorPC (α=0.75) 0.885 0.71 0.801 No

For all three methods, 50 % of the edges present in the gold standard network were randomly selected and assigned a true prior (1 for MEN and BBSR, a random probability in (0.5,1] for PriorPC). For PriorPC, α is given in brackets. MEN and BBSR also use time-series(TS) data. Results are comparable across the three algorithms