makePSN_NamedMatrix: Create patient networks from full matrix of named...
In netDx: Network-based patient classifier

Description Usage Arguments Details Value Examples

Create patient networks from full matrix of named measurements

makePSN_NamedMatrix(
  xpr,
  nm,
  namedSets,
  outDir = tempdir(),
  simMetric = "pearson",
  verbose = TRUE,
  numCores = 1L,
  writeProfiles = TRUE,
  sparsify = FALSE,
  useSparsify2 = FALSE,
  cutoff = 0.3,
  sparsify_edgeMax = Inf,
  sparsify_maxInt = 50,
  minMembers = 1L,
  runSerially = FALSE,
  ...
)

`xpr`	(matrix) rows are measurements, columns are samples. Columns must be named (patient ID)
`nm`	(character) names for measurements corresponding to row order of `xpr`. Must match the names in the named sets specified in `nameSets`
`namedSets`	(list) sets of names to be grouped together. keys are set names, and networks will be named as these. values are character vectors corresponding to groups of names (matching those in `nm`) that are input to network generation
`outDir`	(char) path to directory where networks are written. If missing, is set to tempdir()
`simMetric`	(char) measure of similarity. See `getSimilarity()` for details. If writeProfiles is set to TRUE, must be one of pearson (Pearson correlation) or MI (correlation by mutual information).
`verbose`	(logical) print detailed messages
`numCores`	(integer) number of cores for parallel network generation
`writeProfiles`	(logical) use GeneMANIA's ProfileToNetworkDriver to create interaction networks. If TRUE, this function writes subsets of the original data corresponding to networks to file (profiles). If FALSE, uses getSimilarity() and writes interaction networks.
`sparsify`	(logical). If TRUE, sparsifies patient similarity network. See useSparsify2, sparsify_edgeMax and sparsify_maxInt
`useSparsify2`	(logical). Cleaner sparsification routine. If FALSE, uses new matrix-based sparsify3
`cutoff`	(numeric) patients with similarity smaller than this value are not included in the corresponding interaction network
`sparsify_edgeMax`	(numeric) Max number of edges to include in the final network
`sparsify_maxInt`	(numeric) Max num edges per node in sparsified network.
`minMembers`	(integer) min number of measures in a network for the network to be included. Useful when similarity measures require a minimum number of measures to be meaningful (e.g. minimum of 6 for Pearson correlation)
`runSerially`	(logical) set to TRUE to create nets serially, rather than in parallel
`...`	passed to `getSimilarity()`

Creates patient similarity networks when full matrices of data are provided (e.g. gene expression, questionnaire results). To generate networks from sparse data such as CNVs or indels, use makePSN_RangeSets instead. The rows of the data matrix (xpr) must be named (nm); one network is create for each named set (namedSets). There are two options for the way in which networks are created, depending on the value of writeProfiles. 1. writeProfiles=TRUE: GeneMANIA is used to generate interaction networks and sparsify networks. This only works if the desired measure of similarity is network-level Pearson correlation; an example is networks at the level of pathways. In this case, the user does not explicitly specify a similarity measure and simMetric is ignored. 2. writeProfiles=FALSE: GeneMANIA is not used to generate interaction networks. Rather, netDx uses simMetric to create interaction networks. Networks can be sparsified by excluding weak connections (cutoff).

(char) Basename of files to which networks are written. Side effect of writing interaction networks in outDir

data(xpr,pheno,pathwayList); 
# you may get a warning message that the output directory already
# exists; ignore it
out <- makePSN_NamedMatrix(xpr,rownames(xpr),pathwayList, 
	'.',writeProfiles=TRUE)