Skip to main content

Table 1 Performance on Rosetta data

From: Extracting expression modules from perturbational gene expression compendia

       Top module
GO category # genes method # modules tp R p tp P
conjugation (GO:0000746) 117 ENIGMA 10 43 0.37 3.98E-29 23 0.62
   SAMBA 17 46 0.39 4.10E-29 24 0.55
   ISA 11 42 0.36 1.55E-15 18 0.28
ergosterol biosynthesis (GO:0006696) 26 ENIGMA 4 14 0.54 1.28E-12 9 0.23
   SAMBA 3 16 0.62 1.93E-14 15 0.08
   ISA 1 11 0.42 1.23E-19 11 0.39
cell wall biogenesis (GO:0042546) 32 ENIGMA 1 8 0.25 2.35E-06 8 0.08
   SAMBA 4 9 0.28 6.89E-06 9 0.06
   ISA 1 7 0.22 6.32E-07 7 0.13
iron ion homeostasis (GO:0055072) 38 ENIGMA 4 15 0.39 2.35E-16 11 0.37
   SAMBA 13 16 0.42 3.99E-18 13 0.33
   ISA 2 13 0.34 8.43E-14 13 0.15
oxidative phosphorylation (GO:0006119) 38 ENIGMA 6 23 0.61 3.02E-12 9 0.35
   SAMBA 11 30 0.79 2.34E-32 20 0.44
   ISA 2 8 0.21 1.20E-05 6 0.14
  1. Comparison of ENIGMA, SAMBA and ISA results for selected biological processes targeted by Hughes et al. [4]. The three middle columns give the number of modules enriched for the GO class in the first column, the total number of genes annotated to that GO class in these modules (tp) and the corresponding recall (R). The three last columns contain the enrichment p-value of the top module, the number of true positives (tp) and the proportion of genes in the top module annotated to the respective GO class (precision P).
\