Fig 1.
Examples of between- and within-community Beta distributions.
Four examples of Beta distributions simulated to reflect the distribution function from which either within-community (top panel) or between-community (bottom panel) edge weights that would be sampled in the simulated network.
Fig 2.
Examples of four simulated networks and adjacency matrices.
Four examples of simulated networks of 50 nodes each and four communities. For all simulated networks, the between-group edge weights were sampled from Beta(1, 8). For panels A and C, the within-group edge weights were sampled from Beta(8, 8) while for panels B and D they were sampled from Beta(4, 8).
Table 1.
Logistic kernel: Power and Type I error for cluster evaluation metric simulations.
Table 2.
Linear kernel: Power and Type I error for cluster evaluation metric simulations.
Table 3.
ABIDE dataset subject demographics.
Fig 3.
Distribution of cross correlations in functional connectivity in ABIDE sample.
For subjects with ASD: (A) Average functional connectivity matrix; (B) Standard deviation of the functional connectivity matrix; (C) Cross correlation distribution of functional connectivity. For control subjects: (D) Average functional connectivity matrix; (E) Standard deviation of the functional connectivity matrix; (F) Cross correlation distribution of functional connectivity.
Table 4.
P-values from analysis of ABIDE dataset for binary outcome of diagnostic group.
Table 5.
P-values from analysis of ABIDE dataset for continuous outcome of VIQ.
Table 6.
Analysis of ABIDE dataset: Conventional regression frameworks for optimized algorithmic value.