Skip to main content

Table 7 Simulation I results

From: Integrative network analysis of TCGA data for ovarian cancer

p |E| β N P TPR (skeleton) FDR (skeleton)
50 100 0.5 500 46.0 0.298 (0.410) 0.468 (0.152)
    1000 63.2 0.420 (0.627) 0.333 (0.063)
    2000 78.4 0.600 (0.783) 0.273 (0.032)
   1 500 78.2 0.550 (0.740) 0.294 (0.051)
    1000 92.8 0.676 (0.910) 0.265 (0.019)
    2000 98.4 0.781 (0.960) 0.236 (0.017)
100 200 0.5 500 110.6 0.260 (0.400) 0.528 (0.272)
    1000 124.2 0.328 (0.557) 0.484 (0.104)
    2000 168.4 0.590 (0.825) 0.291 (0.019)
   1 500 163.2 0.539 (0.735) 0.349 (0.098)
    1000 167.8 0.614 (0.892) 0.347 (0.018)
    2000 194.4 0.768 (0.959) 0.216 (0.010)
200 400 0.5 500 252.6 0.225 (0.358) 0.647 (0.444)
    1000 272.8 0.383 (0.597) 0.438 (0.132)
    2000 326.4 0.546 (0.791) 0.337 (0.031)
   1 500 347.2 0.535 (0.825) 0.377 (0.073)
    1000 364.8 0.583 (0.872) 0.359 (0.044)
    2000 396.4 0.698 (0.963) 0.294 (0.028)
  1. Presented in the table are the average number of predicted edges (P), true positive rate (TPR) and false discovery rate (FDR) for both directed and undirected (skeleton) edges over 10 replicated samples in each setting (p |E|, β, N). The number of edges |E| is set to be 2p.