Skip to main content

Table 1 The preset non-zero predictors and their assumed effect values of the different simulation scenarios

From: Gsslasso Cox: a Bayesian hierarchical model for predicting survival and detecting associated genes by incorporating pathway information

Simulation
scenarios
Group, non-zero predictors and effect size
1 non-overlap group
Group group1 group5 group20
predictors {x5 x 20 x40} {x210 x 220 x240} {x975 x995}
2 overlap group
Group group1 group5 group20
predictors {x5 x 20 x40} {x210 x 220 x240} {x975 x995}
3 varying group size (4/20/50)
Group group1 group11
predictors {x1 x 2 x 3 x4} {x501 x 502 x 503 x504}
4 varying number of non-null groups (8/3/1)
Group group1 group2 group7 group8 group11 group12 group19 group20
predictors {x5} {x55} {x305} {x355} {x505} {x555} {x905} {x955}
Group group1 group8 group20
predictors {x5 x 15 x25} {x355 x 365 x375} {x905 x915}
Group group1
predictors {x5 x 10 x 15 x 20 x 25 x 30 x 35 x40}
5 varying correlation within group (r = 0.0/0.5/0.7)
Group group1 group5 group20
predictors {x5 x 20 x40} {x210 x 220 x240} {x975 x995}
Effect size for above simulation scenarios
  0.8 −0.7 1.0 −0.9 −0.8 0.9 −1.0 0.7
6 varying effect size
Group group1 group5 group20
predictors {x5 x 20 x40} {x210 x 220 x240} {x975 x995}
Effect size for scenario 6
  (−2, 2) −0.7 1.0 −0.9 −0.8 0.9 −1.0 0.7
  1. Note:{} quotes the predictors within a group.