Table 1. Summary of simulations.

Item	Small	Medium	Large
Subjects	1000	1000	1000
Features in $\mathcal{X}$	5165	15165	25165
Features in S	50	50	50
Features in X	15	15	15
Features in X′	100	100	100
Noise Features in $\mathcal{X}$	5000	15000	25000
Features in $\mathcal{G}$	5140	10140	20140
Features in H	30	30	30
Features in G	10	10	10
Features in G′	100	100	100
Noise Features in $\mathcal{G}$	5000	10000	20000

The number of features across the simulations for the small, medium, and large number of features are summarized in the table. The same parameters for weak, moderate, and strong signal strengths were generated across each number of features for a total of 9 (3 × 3) different simulations, where each simulation contained 500 replications. The relationships between each of the components within $\mathcal{X}$ and $\mathcal{G}$ are described in Fig 3. In summary, S denotes the features within $\mathcal{X}$ that are related to the outcome but not related to features within $\mathcal{G}$; X denotes the network features within $\mathcal{X}$ that are related to some features within $\mathcal{G}$ and related to the outcome; and X′ denotes the features within $\mathcal{X}$ that are related to some features within $\mathcal{G}$ but not related to the outcome. Similar relationships within $\mathcal{G}$ are described by H, G, and G′.