generateNetwork: Network inference with Gaussian graphical models (GGMs) or...
In krumsiek/MoDentify: Phenotype-driven module identification

Description Usage Arguments Value References Examples

Network inference with Gaussian graphical models (GGMs) or Pearson correlation

1
2
3

generateNetwork(data, covars = NULL, annotations,
  correlation.type = "partial", alpha = 0.05,
  correction.method = "bonferroni")

`data`	a `data.table` or matrix containing the data. The columns correspond to the variables (e.g. metabolites), the rows to the observations.
`covars`	a `data.table` containing covariates to correct for. The columns correspond to the different covariates, the rows to the observations.
`annotations`	a `data.table` containing annotations for the variables (e.g. pathway annotations). The columns correspond to the different annotations, the rows to the variables
`correlation.type`	type of correlation to be estimated. Can either be "pearson", or "partial".
`alpha`	significance level (type 1 error) for multiple testing correction.
`correction.method`	the method that should be used for multiple testing correction ("bonferroni", "BH", "BY", "fdr", "holm", "hochberg", "hommel", "none"). Default is bonferroni. See `p.adjust`.

a network containing the variables as nodes as an igraph object.

\insertRef

Krumsiek2011MoDentify

data(qmdiab.data)
data(qmdiab.annos)

net.graph <- generateNetwork(
  data = qmdiab.data, annotations = qmdiab.annos,
  alpha = 0.05, correction.method = "bonferroni"
)