Skip to main content

Table 2 Performance of OMDR

From: Identification of multiple gene-gene interactions for ordinal phenotypes

  Single-locus classifier Two-locus classifier Three-locus classifier
  EP (SEP) GCVC TSTB TRTB EP GCVC TSTB TRTB EP (TEP) GCVC TSTB TRTB
Model 11 0.38 (0.25) 0.689 0.055 0.097 1.00 1.000 0.444 0.443 1.00 (0.05) 0.514 0.429 0.455
Model 12 0.27 (0.14) 0.667 0.043 0.093 1.00 1.000 0.264 0.278 1.00 (0.05) 0.518 0.256 0.315
Model 13 0.26 (0.16) 0.615 0.046 0.095 0.93 0.883 0.139 0.174 0.87 (0.05) 0.458 0.128 0.224
Model 21 1.00 (0.55) 0.883 0.159 0.175 1.00 1.000 0.395 0.401 1.00 (0.04) 0.480 0.388 0.417
Model 22 0.99 (0.53) 0.881 0.150 0.167 1.00 1.000 0.297 0.306 1.00 (0.05) 0.432 0.267 0.330
Model 23 0.95 (0.54) 0.874 0.115 0.134 0.99 0.993 0.230 0.249 1.00 (0.08) 0.504 0.224 0.292
Model 31 1.00 (1.00) 1.000 0.225 0.229 1.00 1.000 0.514 0.513 1.00 (0.03) 0.507 0.505 0.527
Model 32 1.00 (1.00) 0.997 0.192 0.197 1.00 1.000 0.373 0.375 1.00 (0.06) 0.531 0.356 0.397
Model 33 0.44 (0.31) 0.714 0.061 0.099 1.00 0.997 0.194 0.208 0.99 (0.05) 0.537 0.185 0.245
Model 41 1.00 (0.56) 0.904 0.131 0.144 1.00 0.999 0.252 0.264 1.00 (0.06) 0.544 0.255 0.300
Model 42 1.00 (0.55) 0.887 0.149 0.164 0.90 0.883 0.204 0.217 0.94 (0.05) 0.454 0.189 0.247
Model 43 0.73 (0.41) 0.816 0.083 0.107 0.28 0.485 0.068 0.146 0.23 (0.02) 0.346 0.061 0.197
Model 51 1.00 (0.51) 0.865 0.341 0.355 1.00 1.000 0.478 0.484 1.00 (0.05) 0.480 0.477 0.512
Model 52 1.00 (0.51) 0.885 0.247 0.263 1.00 0.998 0.334 0.344 1.00 (0.05) 0.473 0.325 0.377
Model 53 0.91 (0.46) 0.860 0.092 0.117 0.38 0.557 0.086 0.153 0.47 (0.04) 0.426 0.099 0.204
Average 0.80 (0.50) 0.836 0.139 0.162 0.90 0.920 0.285 0.304 0.90 (0.05) 0.483 0.276 0.336
  1. It is presented with EP (empirical power), average GCVC (K = 1), average TSTB (testing tau-b), average TRTB (training tau-b), and their average values over models. Note that true causal factor was a two-locus classifier (i.e., two-way interaction). SEP indicates EP of single-locus classifier whose EP is largest among all single-locus classifiers included in the true causal interaction. TEP indicates EP of three-locus classifier whose EP is largest among all three-locus classifiers containing the true causal interaction.