gen_graphs_sf: Generate Covariate Dependent Scale-Free Graphs
In berryni/mDINGO: DINGO: Differential Network Analysis in Genomics

Description Arguments Value Author(s)

Use igraph to generate synthetic underlying scale-free graph structure for simulated OTU data. A list of graphs is outputted, the disease, change, control, and global graphs. See “values" for full info. The disease graph is a scale free network generated with Barab\'asi-Albert algorithm. A hub node is then selected and expanded until it is at least as large as 'hubsize' of the full graph. This hub is then permuted to form the control graph.

`numNodes`	The number of nodes to include in created graphs.
`edgePower`	Default value of 1. The power of the preferential attachment. Determines strength of scale-free structure.
`hubgraphchangeprop`	Defaults to .6. Probability of altering each edge in the graph subset that is determined to be the hub (disease generating phenomena).
`nonhubgraphchangeprop`	Defaults to .05. Probability of altering each edge not in the graph subset that is determined to be the hub.
`hubsize`	Defaults to 1/3. Must be between 0 and 1. Dictates size of disease generating hub node. 1/3 indicates that at least 1/3 of the entire graph must be detached from the disease generation node. Larger number -> smaller hub node.

List of 4 igraph objects.

`disease`	Dependence graph for disease having patient subset.
`change`	Graph outlining what changed between disease and control graphs.
`control`	Dependence graph for control patients. Result of applying change graph to disease graph.
`global`	Intersection of disease and control graph edges.