Skip to main content

scstruc: causal assessment of gene regulatory network using Bayesian network

8 Plotting functions

The library has various functions for visualizing the inferred networks with highlighting the important regulatory relationships and comparing the inferred networks by multiple algorithms, using the ggraph, tidygraph and ggfx. These can be useful for the diagnosis and for the publication figure. Mostly, the function accepts the bn object from bnlearn.

8.1 plotNet

plotNet function accepts bn object.

library(scstruc)
sce <- mockSCE()
sce <- logNormCounts(sce)
included_genes <- sample(row.names(sce), 50)
gs <- scstruc(sce, included_genes, changeSymbol=FALSE,
    algorithm="glmnet_BIC", returnData=TRUE)

## Vanilla plot
plotNet(gs$net)

Optionally, data can be passed to fit the parameters and the function colors the edge using the fitted parameters. The nodes will be sized by their degrees. degreeMode argument can be specified for how the degree will be calculated. By default, mode="all" is specified.

8.2 plotAVN

This function is intended specifically for visualizing the bootstrapped networks. You can pass the results of functions returning bn.strength object, to this function. The threshold will be automatically determined if not specified. In the plot, the strength are shown as edge width, and the direction are shown as edge color. The node sizes will be determined by the degree.

library(ggraph)
gs.boot <- scstruc(sce, included_genes, changeSymbol=FALSE, boot=TRUE, R=10,
             algorithm="glmnet_BIC", returnData=TRUE)
#> Bootstrapping specified

plotAVN(gs.boot$net)

Edges can be highlighted by specifying highlightEdges argument, as same as plotNet. In the plotAVN, the with_outer_glow function in ggfx package is used to highlight edges as to preserve the strength and direction mapping in the plot.

8.3 ggraph.compare

Similar to graphviz.compare function in bnlearn, the diagnostic plot can be made by ggraph.compare. This accepts the list of bn object and plot the nodes and each edge associated with the graphs by geom_edge_parallel. The agreement of edges in multiple graphs can be understood with the function, however, it will get too complicated if many graphs are provided.

ggraph.compare(list("boot"=averaged.network(gs.boot$net), "raw"=gs$net))