Package 'multiWGCNA' reference manual

Title:	multiWGCNA
Description:	An R package for deeping mining gene co-expression networks in multi-trait expression data. Provides functions for analyzing, comparing, and visualizing WGCNA networks across conditions. multiWGCNA was designed to handle the common case where there are multiple biologically meaningful sample traits, such as disease vs wildtype across development or anatomical region.
Authors:	Dario Tommasini [aut, cre] , Brent Fogel [aut, ctb]
Maintainer:	Dario Tommasini <[email protected]>
License:	GPL-3
Version:	1.3.1
Built:	2024-07-04 05:22:58 UTC
Source:	https://github.com/bioc/multiWGCNA

Best matching modules

Description

Find all the modules from dataset1 that have a best match to a module in dataset2 if that module in dataset2 is also a best match to the module in dataset1

Usage

bidirectionalBestMatches(comparisonList, plot = TRUE)
bidirectionalBestMatches(comparisonList, plot = TRUE)

Arguments

`comparisonList`	a list with an elemnt "overlap", which is a data.frame resulting from a call to computeOverlapsFromWGCNA
`plot`	whether to generate a heatmap; default is TRUE

Value

A ggplot object

Author(s)

Dario Tommasini

Examples

library(ExperimentHub)
eh = ExperimentHub()
eh_query = query(eh, c("multiWGCNAdata"))
astrocyte_networks = eh_query[["EH8222"]]
comparisonList = list()
comparisonList$overlaps = computeOverlapsFromWGCNA(astrocyte_networks$EAE, astrocyte_networks$WT)
bidirectionalBestMatches(comparisonList)

library(ExperimentHub)
eh = ExperimentHub()
eh_query = query(eh, c("multiWGCNAdata"))
astrocyte_networks = eh_query[["EH8222"]]
comparisonList = list()
comparisonList$overlaps = computeOverlapsFromWGCNA(astrocyte_networks$EAE, astrocyte_networks$WT)
bidirectionalBestMatches(comparisonList)

cleanDatExpr

Description

A function that converts a data.frame where row 1 is gene symbols to a numeric matrix where columns are genes and rows are samples for compatibility with most WGCNA functions.

Usage

cleanDatExpr(datExpr, checkGenesSamples = FALSE)
cleanDatExpr(datExpr, checkGenesSamples = FALSE)

Arguments

`datExpr`	a data.frame were columns are samples and rows are samples and the gene symbols are in the first row
`checkGenesSamples`	call the WGCNA function checkGenesSamples?

Value

Returns a datExpr with rows as samples and columns as genes

Author(s)

Dario Tommasini

Examples

library(ExperimentHub)
eh = ExperimentHub()
eh_query = query(eh, c("multiWGCNAdata"))
astrocyte_se = eh_query[["EH8223"]]
datExpr = data.frame(X = rownames(assays(astrocyte_se)[[1]]), assays(astrocyte_se)[[1]])
cleanDatExpr(datExpr)

library(ExperimentHub)
eh = ExperimentHub()
eh_query = query(eh, c("multiWGCNAdata"))
astrocyte_se = eh_query[["EH8223"]]
datExpr = data.frame(X = rownames(assays(astrocyte_se)[[1]]), assays(astrocyte_se)[[1]])
cleanDatExpr(datExpr)

Coexpression Line Graph

Description

Plots a line graph showing the co-expression of selected genes across samples

Usage

coexpressionLineGraph(datExpr, splitBy = 1, fontSize = 2.15, colors = NULL)
coexpressionLineGraph(datExpr, splitBy = 1, fontSize = 2.15, colors = NULL)

Arguments

`datExpr`	a data.frame with genes as rows and samples as columns
`splitBy`	how much to split genes by on line graph
`fontSize`	the font size of the gene labels
`colors`	a vector of colors; default is random colors generated by colors function

Value

a ggplot object

Author(s)

Dario Tommasini

Examples

library(ExperimentHub)
eh = ExperimentHub()
eh_query = query(eh, c("multiWGCNAdata"))
astrocyte_networks = eh_query[["EH8222"]]
datExpr = GetDatExpr(astrocyte_networks[[1]], 
  genes = topNGenes(astrocyte_networks$EAE, "EAE_015", 20))
coexpressionLineGraph(datExpr)	+ 
  geom_vline(xintercept = 20.5, linetype='dashed')

library(ExperimentHub)
eh = ExperimentHub()
eh_query = query(eh, c("multiWGCNAdata"))
astrocyte_networks = eh_query[["EH8222"]]
datExpr = GetDatExpr(astrocyte_networks[[1]], 
  genes = topNGenes(astrocyte_networks$EAE, "EAE_015", 20))
coexpressionLineGraph(datExpr)	+ 
  geom_vline(xintercept = 20.5, linetype='dashed')

computeOverlapsFromWGCNA

Description

Computes overlap between the modules of two objects of class WGCNA

Usage

computeOverlapsFromWGCNA(dataset1, dataset2)
computeOverlapsFromWGCNA(dataset1, dataset2)

Arguments

`dataset1`	an object of class WGCNA to compare with dataset2
`dataset2`	an object of class WGCNA to compare with dataset1

Value

Returns a data.frame showing the overlap results for modules from dataset1 with dataset2

Author(s)

Dario Tommasini

Examples

library(ExperimentHub)
eh = ExperimentHub()
eh_query = query(eh, c("multiWGCNAdata"))
astrocyte_networks = eh_query[["EH8222"]]
computeOverlapsFromWGCNA(astrocyte_networks$EAE, astrocyte_networks$WT)

library(ExperimentHub)
eh = ExperimentHub()
eh_query = query(eh, c("multiWGCNAdata"))
astrocyte_networks = eh_query[["EH8222"]]
computeOverlapsFromWGCNA(astrocyte_networks$EAE, astrocyte_networks$WT)

constructNetworks: Construct all the weighted gene correlation networks

Description

A high level function that returns all networks possible for a given experimental design

Usage

constructNetworks(
  datExpr,
  sampleTable,
  conditions1,
  conditions2,
  write = FALSE,
  alphaLevel = 0.05,
  plot = FALSE,
  ...
)
constructNetworks(
  datExpr,
  sampleTable,
  conditions1,
  conditions2,
  write = FALSE,
  alphaLevel = 0.05,
  plot = FALSE,
  ...
)

Arguments

`datExpr`	either a SummarizedExperiment object or data.frame with genes are rows and samples as columns
`sampleTable`	data.frame with sample names in first column and sample traits in the second and third column. First column should be called "Sample"
`conditions1`	first design conditions, ie healthy/disease
`conditions2`	second design conditions, ie frontal lobe/temporal lobe
`write`	write results out to files?
`alphaLevel`	significance value passed to findBestTrait function, default is 0.05
`plot`	plot modules? Default is false
`...`	Arguments to pass to blockwiseModules function

Value

A list of WGCNA objects, ie level one, two, and three networks.

Author(s)

Dario Tommasini

Examples

library(ExperimentHub)
eh = ExperimentHub()
eh_query = query(eh, c("multiWGCNAdata"))
autism_se = eh_query[["EH8219"]]
set.seed(1)
autism_se = autism_se[sample(rownames(autism_se), 500),]
sampleTable = colData(autism_se)
conditions1 = unique(sampleTable[,2])
conditions2 = unique(sampleTable[,3])
autism_networks = constructNetworks(autism_se, sampleTable, conditions1[[1]], conditions2[[1]], 
  networkType = "signed", TOMType = "unsigned", 
  power = 10, minModuleSize = 100, maxBlockSize = 25000,
  reassignThreshold = 0, minKMEtoStay = 0, mergeCutHeight = 0,
  numericLabels = TRUE, pamRespectsDendro = FALSE, 
  deepSplit = 4, verbose = 3)
autism_networks[["combined"]]

library(ExperimentHub)
eh = ExperimentHub()
eh_query = query(eh, c("multiWGCNAdata"))
autism_se = eh_query[["EH8219"]]
set.seed(1)
autism_se = autism_se[sample(rownames(autism_se), 500),]
sampleTable = colData(autism_se)
conditions1 = unique(sampleTable[,2])
conditions2 = unique(sampleTable[,3])
autism_networks = constructNetworks(autism_se, sampleTable, conditions1[[1]], conditions2[[1]], 
  networkType = "signed", TOMType = "unsigned", 
  power = 10, minModuleSize = 100, maxBlockSize = 25000,
  reassignThreshold = 0, minKMEtoStay = 0, mergeCutHeight = 0,
  numericLabels = TRUE, pamRespectsDendro = FALSE, 
  deepSplit = 4, verbose = 3)
autism_networks[["combined"]]

Differential co-expresison analysis

Description

Performs a differential co-expression ananlysis given an expression data.frame and a conditions vector

Usage

diffCoexpression(
  datExpr,
  conditions,
  geneList = NULL,
  plot = FALSE,
  method = c("pearson", "spearman"),
  removeFreeNodes = TRUE,
  labelSize = 0.5,
  labelDist = 0,
  shape = "circle",
  degreeForSize = FALSE,
  label = FALSE,
  onlyPositive = FALSE,
  z.threshold = NULL,
  FDR.threshold = 0.05,
  nodeSize = 3
)
diffCoexpression(
  datExpr,
  conditions,
  geneList = NULL,
  plot = FALSE,
  method = c("pearson", "spearman"),
  removeFreeNodes = TRUE,
  labelSize = 0.5,
  labelDist = 0,
  shape = "circle",
  degreeForSize = FALSE,
  label = FALSE,
  onlyPositive = FALSE,
  z.threshold = NULL,
  FDR.threshold = 0.05,
  nodeSize = 3
)

Arguments

`datExpr`	a data.frame containing expression values
`conditions`	a vector containing conditions for the samples
`geneList`	vector of genes, will use all genes if NULL (default)
`plot`	plot a network?
`method`	either "pearson" or "spearman"
`removeFreeNodes`	remove free nodes from network?
`labelSize`	label size
`labelDist`	distance from labels to nodes
`shape`	shape of nodes
`degreeForSize`	should node size correspond to degree?
`label`	label nodes?
`onlyPositive`	only draw positive correlations?
`z.threshold`	z-score threshold
`FDR.threshold`	FDR threshold
`nodeSize`	size of node

Value

A list including a matrix of z-scores, a matrix of raw p-values, a matrix of adjusted p-values, and a summary data.frame

Author(s)

Dario Tommasini

Examples

library(ExperimentHub)
eh = ExperimentHub()
eh_query = query(eh, c("multiWGCNAdata"))
astrocyte_se = eh_query[["EH8223"]]
datExpr = assays(astrocyte_se)[[1]]
diffCoexpression(datExpr, c(rep(1,20), rep(2,16)), 
  geneList = c("Gfap", "Vim", "Aspg", "Serpina3n", "Cp", "Osmr", "Cd44", 
    "Cxcl10", "Hspb1", "Timp1", "S1pr3", "Steap4", "Lcn2"))

library(ExperimentHub)
eh = ExperimentHub()
eh_query = query(eh, c("multiWGCNAdata"))
astrocyte_se = eh_query[["EH8223"]]
datExpr = assays(astrocyte_se)[[1]]
diffCoexpression(datExpr, c(rep(1,20), rep(2,16)), 
  geneList = c("Gfap", "Vim", "Aspg", "Serpina3n", "Cp", "Osmr", "Cd44", 
    "Cxcl10", "Hspb1", "Timp1", "S1pr3", "Steap4", "Lcn2"))

Differential module expression

Description

Runs (and plots) the differential module expression analysis

Usage

diffModuleExpression(
  WGCNAobject,
  geneList,
  design,
  plotTitle = NULL,
  mode = c("PC1", "Zscore"),
  testColumn = 2,
  refColumn = 3,
  test = c("ANOVA", "PERMANOVA"),
  plot = TRUE
)
diffModuleExpression(
  WGCNAobject,
  geneList,
  design,
  plotTitle = NULL,
  mode = c("PC1", "Zscore"),
  testColumn = 2,
  refColumn = 3,
  test = c("ANOVA", "PERMANOVA"),
  plot = TRUE
)

Arguments

`WGCNAobject`	WGCNA object
`geneList`	vector of genes in WGCNAobject
`design`	the sampleTable
`plotTitle`	title for the plot
`mode`	either PC1 or Zscore, default is PC1
`testColumn`	the column of the sampleTable to be resolved
`refColumn`	the column of the sampleTable to be used as biological variation
`test`	statistical test to perform, either "ANOVA" or "PERMANOVA"
`plot`	generate a plot?

Value

a data.frame with the resulting p-values

Examples

library(ExperimentHub)
eh = ExperimentHub()
eh_query = query(eh, c("multiWGCNAdata"))
astrocyte_se = eh_query[["EH8223"]]
sampleTable = colData(astrocyte_se)
astrocyte_networks = eh_query[["EH8222"]]
diffModuleExpression(astrocyte_networks[["combined"]], 
  topNGenes(astrocyte_networks$combined, "combined_013"), 
  sampleTable,
  test = "ANOVA",
  plotTitle = "combined_013",
  plot = TRUE)

library(ExperimentHub)
eh = ExperimentHub()
eh_query = query(eh, c("multiWGCNAdata"))
astrocyte_se = eh_query[["EH8223"]]
sampleTable = colData(astrocyte_se)
astrocyte_networks = eh_query[["EH8222"]]
diffModuleExpression(astrocyte_networks[["combined"]], 
  topNGenes(astrocyte_networks$combined, "combined_013"), 
  sampleTable,
  test = "ANOVA",
  plotTitle = "combined_013",
  plot = TRUE)

Draw multiWGCNA network

Description

Draw a network where nodes are modules and edges represent significant gene overlap. Modules are sorted by levels 1, 2, and 3.

Usage

drawMultiWGCNAnetwork(
  WGCNAlist,
  comparisonList,
  moduleOfInterest,
  design,
  overlapCutoff = 0,
  padjCutoff = 1,
  removeOutliers = TRUE,
  alpha = 1e-50,
  layout = NULL,
  hjust = 0.4,
  vjust = 0.3,
  width = 0.5,
  colors = NULL
)
drawMultiWGCNAnetwork(
  WGCNAlist,
  comparisonList,
  moduleOfInterest,
  design,
  overlapCutoff = 0,
  padjCutoff = 1,
  removeOutliers = TRUE,
  alpha = 1e-50,
  layout = NULL,
  hjust = 0.4,
  vjust = 0.3,
  width = 0.5,
  colors = NULL
)

Arguments

`WGCNAlist`	list of WGCNA objects
`comparisonList`	the list of overlap comparisons ie from iterate(myNetworks, overlapComparisons, ...)
`moduleOfInterest`	module of interest, ie "combined_001"
`design`	the sampleTable design matrix
`overlapCutoff`	cutoff to remove module correspondences with less than this number of genes
`padjCutoff`	cutoff to remove module correspondences above this significance value
`removeOutliers`	remove outlier modules?
`alpha`	alpha level of significance
`layout`	layout of network to be passed to plot function of igraph object, defaults to multiWGCNA custom layout
`hjust`	horizontal justification of labels
`vjust`	vertical justification of labels
`width`	width of labels
`colors`	colors to use for modules, should be the same length as the number of WGCNA objects in the WGCNAlist. Defaults to random colors for each condition.

Value

an igraph plot

Author(s)

Dario Tommasini

Examples

library(ExperimentHub)
eh = ExperimentHub()
eh_query = query(eh, c("multiWGCNAdata"))
astrocyte_se = eh_query[["EH8223"]]
sampleTable = colData(astrocyte_se)
astrocyte_networks = eh_query[["EH8222"]]
results = list()
results$overlaps = iterate(astrocyte_networks, overlapComparisons, plot=FALSE)
drawMultiWGCNAnetwork(astrocyte_networks, 
  results$overlaps, 
  "combined_013", 
  sampleTable)
  
library(ExperimentHub)
eh = ExperimentHub()
eh_query = query(eh, c("multiWGCNAdata"))
astrocyte_se = eh_query[["EH8223"]]
sampleTable = colData(astrocyte_se)
astrocyte_networks = eh_query[["EH8222"]]
results = list()
results$overlaps = iterate(astrocyte_networks, overlapComparisons, plot=FALSE)
drawMultiWGCNAnetwork(astrocyte_networks, 
  results$overlaps, 
  "combined_013", 
  sampleTable)

Get expression data

Description

Returns the expression data frame a WGCNA object as a data.frame

Usage

GetDatExpr(object, genes = NULL)
GetDatExpr(object, genes = NULL)

Arguments

`object`	An object of class WGCNA
`genes`	a list of genes to subset to; default is NULL

Value

a data.frame

Author(s)

Dario Tommasini

Examples

library(ExperimentHub)
eh = ExperimentHub()
eh_query = query(eh, c("multiWGCNAdata"))
astrocyte_networks = eh_query[["EH8222"]]
datExpr = GetDatExpr(astrocyte_networks[[1]], 
  genes = topNGenes(astrocyte_networks$EAE, "EAE_015", 20))
coexpressionLineGraph(datExpr)	+ 
  geom_vline(xintercept = 20.5, linetype='dashed')

library(ExperimentHub)
eh = ExperimentHub()
eh_query = query(eh, c("multiWGCNAdata"))
astrocyte_networks = eh_query[["EH8222"]]
datExpr = GetDatExpr(astrocyte_networks[[1]], 
  genes = topNGenes(astrocyte_networks$EAE, "EAE_015", 20))
coexpressionLineGraph(datExpr)	+ 
  geom_vline(xintercept = 20.5, linetype='dashed')

getPreservation

Description

Performs a network preservation analysis

Usage

getPreservation(reference, test, nPermutations = 100, write = FALSE)
getPreservation(reference, test, nPermutations = 100, write = FALSE)

Arguments

`reference`	reference network of class WGCNA
`test`	test network of class WGCNA
`nPermutations`	number of permutations to perform; at least 50 permutations
`write`	write to file?

Value

a data.frame summarizing results of preservation analysis

Author(s)

Dario Tommasini

iterate: Iterate function across networks

Description

A high level function that iterates functions across a list of WGCNA objects

Usage

iterate(WGCNAlist, FUN, ...)
iterate(WGCNAlist, FUN, ...)

Arguments

`WGCNAlist`	a vector of objects of type WGCNAobject
`FUN`	function to iterate, either overlapComparisons or preservationComparisons
`...`	argmuents to be passed on to overlapComparisons or preservationComparisons

Value

a comparison list from overlapComparisons or preservationComparisons

Author(s)

Dario Tommasini

Examples

library(ExperimentHub)
eh = ExperimentHub()
eh_query = query(eh, c("multiWGCNAdata"))
astrocyte_networks = eh_query[["EH8222"]]
results = list()
iterate(astrocyte_networks, overlapComparisons, plot=FALSE)

library(ExperimentHub)
eh = ExperimentHub()
eh_query = query(eh, c("multiWGCNAdata"))
astrocyte_networks = eh_query[["EH8222"]]
results = list()
iterate(astrocyte_networks, overlapComparisons, plot=FALSE)

Generate a trait table from a sample table

Description

Generates a WGCNA-compatible trait table from a sampleTable dataframe

Usage

makeTraitTable(inputTable, column, detectNumbers = FALSE)
makeTraitTable(inputTable, column, detectNumbers = FALSE)

Arguments

`inputTable`	the sampleTable data.frame
`column`	the column from the sampleTable to use as traits
`detectNumbers`	whether to consider traits with numbers as numerical rather than categorical variables

Value

a data.frame with integer values denoting the categorical sample traits

Examples

sampleTable = data.frame(Sample = c(paste0("EAE", 1:10), paste0("WT", 1:10)), 
                         Disease = c(rep("EAE", 10), rep("WT", 10)), 
                         Region = c(rep(c("Cbl", "Sc"), 5))) 
makeTraitTable(sampleTable, 2)

sampleTable = data.frame(Sample = c(paste0("EAE", 1:10), paste0("WT", 1:10)), 
                         Disease = c(rep("EAE", 10), rep("WT", 10)), 
                         Region = c(rep(c("Cbl", "Sc"), 5))) 
makeTraitTable(sampleTable, 2)

Module comparison plot

Description

A plotting function that returns a heatmap and barplot for a module

Usage

moduleComparisonPlot(overlapDf, dataset1, dataset2)
moduleComparisonPlot(overlapDf, dataset1, dataset2)

Arguments

`overlapDf`	a data.frame resulting from a call to computeOverlapsFromWGCNA
`dataset1`	an object of class WGCNA to compare with dataset2
`dataset2`	an object of class WGCNA to compare with dataset1

Value

Returns a ggplot object (flowplot and heatmap) showing the module correspondence between two objects of class WGCNA

Author(s)

Dario Tommasini

Examples

library(ExperimentHub)
eh = ExperimentHub()
eh_query = query(eh, c("multiWGCNAdata"))
astrocyte_networks = eh_query[["EH8222"]]
overlapDf = computeOverlapsFromWGCNA(astrocyte_networks$EAE, astrocyte_networks$WT)
moduleComparisonPlot(overlapDf, astrocyte_networks$EAE, astrocyte_networks$WT)

library(ExperimentHub)
eh = ExperimentHub()
eh_query = query(eh, c("multiWGCNAdata"))
astrocyte_networks = eh_query[["EH8222"]]
overlapDf = computeOverlapsFromWGCNA(astrocyte_networks$EAE, astrocyte_networks$WT)
moduleComparisonPlot(overlapDf, astrocyte_networks$EAE, astrocyte_networks$WT)

Plots an expression profile for a module

Description

A plotting function that returns a heatmap and barplot for a module

Usage

moduleExpressionPlot(
  WGCNAobject,
  geneList,
  mode = c("PC1", "averageZscore"),
  legend = FALSE,
  title = NULL,
  clusterGenes = FALSE
)
moduleExpressionPlot(
  WGCNAobject,
  geneList,
  mode = c("PC1", "averageZscore"),
  legend = FALSE,
  title = NULL,
  clusterGenes = FALSE
)

Arguments

`WGCNAobject`	an object of class WGCNAobject
`geneList`	a vector of gene names to be extracted from WGCNAobject
`mode`	use first principal component or averageZscore?
`legend`	plot legend?
`title`	title of the plot
`clusterGenes`	cluster heatmap genes by hierarchical clustering?

Value

a patchworked ggplot object

Author(s)

Dario Tommasini

Examples

library(ExperimentHub)
eh = ExperimentHub()
eh_query = query(eh, c("multiWGCNAdata"))
astrocyte_networks = eh_query[["EH8222"]]
moduleExpressionPlot(astrocyte_networks[["combined"]], 
  geneList = topNGenes(astrocyte_networks$combined, "combined_013"))

library(ExperimentHub)
eh = ExperimentHub()
eh_query = query(eh, c("multiWGCNAdata"))
astrocyte_networks = eh_query[["EH8222"]]
moduleExpressionPlot(astrocyte_networks[["combined"]], 
  geneList = topNGenes(astrocyte_networks$combined, "combined_013"))

Module to module heatmap

Description

Returns a heatmap where color corresponds to FDR-adjusted overlap (hypergeometric test) and the label corresponds to the number of overlapping genes

Usage

moduleToModuleHeatmap(
  comparisonDf,
  dataset1 = NULL,
  dataset2 = NULL,
  trait1 = NULL,
  trait2 = NULL,
  list1 = NULL,
  list2 = NULL,
  filterByTrait = FALSE,
  alphaLevel = 0.05
)
moduleToModuleHeatmap(
  comparisonDf,
  dataset1 = NULL,
  dataset2 = NULL,
  trait1 = NULL,
  trait2 = NULL,
  list1 = NULL,
  list2 = NULL,
  filterByTrait = FALSE,
  alphaLevel = 0.05
)

Arguments

`comparisonDf`	the data.frame output of computeOverlapFromWGCNA
`dataset1`	optional; WGCNA object for dataset 1
`dataset2`	optional; WGCNA object for dataset 2
`trait1`	optional; subset to modules correlated to this trait for dataset 1
`trait2`	optional; subset to modules correlated to this trait for dataset 2
`list1`	subset to this list of modules for dataset 1
`list2`	subset to this list of modules for dataset 2
`filterByTrait`	only plot for modules that correlate with some trait?
`alphaLevel`	the alpha level of significance for module-trait correlation, defaults to 0.05

Value

A ggplot object

Examples

library(ExperimentHub)
eh = ExperimentHub()
eh_query = query(eh, c("multiWGCNAdata"))
astrocyte_networks = eh_query[["EH8222"]]
overlapDf = computeOverlapsFromWGCNA(astrocyte_networks$EAE, astrocyte_networks$WT)
moduleToModuleHeatmap(overlapDf)

library(ExperimentHub)
eh = ExperimentHub()
eh_query = query(eh, c("multiWGCNAdata"))
astrocyte_networks = eh_query[["EH8222"]]
overlapDf = computeOverlapsFromWGCNA(astrocyte_networks$EAE, astrocyte_networks$WT)
moduleToModuleHeatmap(overlapDf)

name: Name of WGCNAobject

Description

Returns the name of a WGCNAobject.

Usage

name(WGCNAobject)
name(WGCNAobject)

Arguments

WGCNAobject

an object of class WGCNA

Value

Returns the name of the WGCNA object, ie "EAE" for astrocyte_networks$EAE.

Examples

library(ExperimentHub)
eh = ExperimentHub()
eh_query = query(eh, c("multiWGCNAdata"))
astrocyte_networks = eh_query[["EH8222"]]
name(astrocyte_networks$EAE)
library(ExperimentHub)
eh = ExperimentHub()
eh_query = query(eh, c("multiWGCNAdata"))
astrocyte_networks = eh_query[["EH8222"]]
name(astrocyte_networks$EAE)

Overlap comparisons

Description

Compares modules between two objects of type WGCNAobjects within a WGCNAobject list given the indices. Recommended to be used in conjunction with the iterate function.

Usage

overlapComparisons(
  comparisonList,
  WGCNAlist,
  first,
  second,
  element,
  plot = TRUE,
  write = FALSE
)
overlapComparisons(
  comparisonList,
  WGCNAlist,
  first,
  second,
  element,
  plot = TRUE,
  write = FALSE
)

Arguments

`comparisonList`	a list passed by the iterate function
`WGCNAlist`	list of objects of class WGCNA
`first`	index of first WGCNA object
`second`	index of second WGCNA object
`element`	element position in the comparison list (passed by iterate function)
`plot`	generate plots?
`write`	write results to file?

Value

A list, in which the first element is a data.frame showing the overlap results and the second element is a data.frame showing the best matching modules between the two WGCNA objects.

Author(s)

Dario Tommasini

Examples

library(ExperimentHub)
eh = ExperimentHub()
eh_query = query(eh, c("multiWGCNAdata"))
astrocyte_networks = eh_query[["EH8222"]]
results = list()
results$overlaps = iterate(astrocyte_networks, overlapComparisons, plot=FALSE)

library(ExperimentHub)
eh = ExperimentHub()
eh_query = query(eh, c("multiWGCNAdata"))
astrocyte_networks = eh_query[["EH8222"]]
results = list()
results$overlaps = iterate(astrocyte_networks, overlapComparisons, plot=FALSE)

Perform ANOVA

Description

Test association between module expression to traits using ANOVA

Usage

performANOVA(datExpr, design, testCondition, refCondition, alphaLevel = 0.05)
performANOVA(datExpr, design, testCondition, refCondition, alphaLevel = 0.05)

Arguments

`datExpr`	expression data.frame
`design`	the sampleTable
`testCondition`	test column in sampleTable
`refCondition`	reference column in sampleTable
`alphaLevel`	the significance level

Value

a data.frame with p-values for each association

Permutation test results

Description

The results of running the PreservationPermutationTest in the astrocyte vignette. This is provided since this function is quite slow. Please see the astrocyte vignette for more details.

Usage

data(permutationTestResults)
data(permutationTestResults)

Format

A list of data.frames containing preservation results for each permutation

Preservation Comparison Scatterplot

Description

A plotting function that draws a scatterplot of preservation scores between two WGCNA objects

Usage

preservationComparisonPlot(
  preservationList,
  dataset1,
  dataset2,
  alphaLevel = 0.05,
  outliers = FALSE
)
preservationComparisonPlot(
  preservationList,
  dataset1,
  dataset2,
  alphaLevel = 0.05,
  outliers = FALSE
)

Arguments

`preservationList`	a list resulting from a call to preservationComparisons
`dataset1`	an object of class WGCNAobject to compare with dataset2
`dataset2`	an object of class WGCNAobject to compare with dataset1
`alphaLevel`	alpha level of significance, default is 0.05
`outliers`	leave outlier modules? By default these are removed

Value

a ggplot object

Author(s)

Dario Tommasini

Examples

library(ExperimentHub)
eh = ExperimentHub()
eh_query = query(eh, c("multiWGCNAdata"))
astrocyte_networks = eh_query[["EH8222"]]
results = list()
results$preservation=iterate(astrocyte_networks[c("EAE", "WT")], 
  preservationComparisons, 
  write=FALSE, 
  plot=FALSE, 
  nPermutations=2)
preservationComparisonPlot(results$preservation$EAE_vs_WT, 
  astrocyte_networks$EAE, 
  astrocyte_networks$WT)
  
library(ExperimentHub)
eh = ExperimentHub()
eh_query = query(eh, c("multiWGCNAdata"))
astrocyte_networks = eh_query[["EH8222"]]
results = list()
results$preservation=iterate(astrocyte_networks[c("EAE", "WT")], 
  preservationComparisons, 
  write=FALSE, 
  plot=FALSE, 
  nPermutations=2)
preservationComparisonPlot(results$preservation$EAE_vs_WT, 
  astrocyte_networks$EAE, 
  astrocyte_networks$WT)

Preservation comparisons

Description

A high level function that performs a perservation comparison between two WGCNAobjects in a WGCNAlist, usually supplied by iterate function

Usage

preservationComparisons(
  comparisonList,
  WGCNAlist,
  first,
  second,
  element,
  plot = FALSE,
  write = FALSE,
  alphaLevel = 0.05,
  nPermutations = 100
)
preservationComparisons(
  comparisonList,
  WGCNAlist,
  first,
  second,
  element,
  plot = FALSE,
  write = FALSE,
  alphaLevel = 0.05,
  nPermutations = 100
)

Arguments

`comparisonList`	a list passed by the iterate function
`WGCNAlist`	list of objects of type WGCNAobject
`first`	index of first WGCNAobject
`second`	index of second WGCNAobject
`element`	element position in the comparison list (passed by iterate function)
`plot`	generate plots?
`write`	write results to file?
`alphaLevel`	alpha level of significance for module-trait correlation
`nPermutations`	number of permutations, defaults to 100

Value

a list of preservation comparisons results across levels 1, 2, 3

Author(s)

Dario Tommasini

Examples

library(ExperimentHub)
eh = ExperimentHub()
eh_query = query(eh, c("multiWGCNAdata"))
astrocyte_networks = eh_query[["EH8222"]]
results = list()
iterate(astrocyte_networks[c("EAE", "WT")], 
  preservationComparisons, 
  write=FALSE, 
  plot=FALSE, 
  nPermutations=2)

library(ExperimentHub)
eh = ExperimentHub()
eh_query = query(eh, c("multiWGCNAdata"))
astrocyte_networks = eh_query[["EH8222"]]
results = list()
iterate(astrocyte_networks[c("EAE", "WT")], 
  preservationComparisons, 
  write=FALSE, 
  plot=FALSE, 
  nPermutations=2)

Performs a permutation test to determine if a null distribution of expected preservation scores for modules in this dataset if the labels were randomly assigned. Please look at the astrocyte vignette for more info.

Usage

PreservationPermutationTest(
  referenceDatExpr,
  design,
  constructNetworksIn,
  testPreservationIn,
  nPermutations = 100,
  nPresPermutations = 100,
  ...
)
PreservationPermutationTest(
  referenceDatExpr,
  design,
  constructNetworksIn,
  testPreservationIn,
  nPermutations = 100,
  nPresPermutations = 100,
  ...
)

Arguments

`referenceDatExpr`	the combined datExpr
`design`	the sampleTable
`constructNetworksIn`	the condition to use for network construction, e.g. for the astrocyte data, this is "EAE"
`testPreservationIn`	the condition to use for testing preservation, e.g. for the astrocyte data, this was "WT"
`nPermutations`	the number of permutations to perform for permutation test
`nPresPermutations`	the number of permutations to perform in modulePreservation function
`...`	arguments to pass to blockwiseModules function for network construction (should be the same as used for constructing the original network)

Value

A list of data.frames with preservation results for each permutation

Author(s)

Dario Tommasini

Examples

## Not run: 
library(ExperimentHub)
eh = ExperimentHub()
eh_query = query(eh, c("multiWGCNAdata"))
astrocyte_networks = eh_query[["EH8222"]] 
astrocyte_se = eh_query[["EH8223"]] 
sampleTable = colData(astrocyte_se)
results = list()
results$permutation.test = PreservationPermutationTest(astrocyte_networks$combined@datExpr[sample(17000,3000),], 
                                                       sampleTable, 
                                                       constructNetworksIn = "EAE", # Construct networks using EAE samples
                                                       testPreservationIn = "WT", # Test preservation of disease samples in WT samples
                                                       nPermutations = 10, # Number of permutations for permutation test
                                                       nPresPermutations = 10, # Number of permutations for modulePreservation function
                                                       networkType = "signed", TOMType = "unsigned", 
                                                       power = 12, minModuleSize = 100, maxBlockSize = 25000,
                                                       reassignThreshold = 0, minKMEtoStay = 0, mergeCutHeight = 0,
                                                       numericLabels = TRUE, pamRespectsDendro = FALSE, 
                                                       deepSplit = 4, verbose = 3
                                                       )

## End(Not run)
## Not run: 
library(ExperimentHub)
eh = ExperimentHub()
eh_query = query(eh, c("multiWGCNAdata"))
astrocyte_networks = eh_query[["EH8222"]] 
astrocyte_se = eh_query[["EH8223"]] 
sampleTable = colData(astrocyte_se)
results = list()
results$permutation.test = PreservationPermutationTest(astrocyte_networks$combined@datExpr[sample(17000,3000),], 
                                                       sampleTable, 
                                                       constructNetworksIn = "EAE", # Construct networks using EAE samples
                                                       testPreservationIn = "WT", # Test preservation of disease samples in WT samples
                                                       nPermutations = 10, # Number of permutations for permutation test
                                                       nPresPermutations = 10, # Number of permutations for modulePreservation function
                                                       networkType = "signed", TOMType = "unsigned", 
                                                       power = 12, minModuleSize = 100, maxBlockSize = 25000,
                                                       reassignThreshold = 0, minKMEtoStay = 0, mergeCutHeight = 0,
                                                       numericLabels = TRUE, pamRespectsDendro = FALSE, 
                                                       deepSplit = 4, verbose = 3
                                                       )

## End(Not run)

PreservationScoreDistribution

Description

Extracts the preservation score distribution from the results of PreservationPermutationTest.

Usage

PreservationScoreDistribution(preservationData, moduleOfInterestSize)
PreservationScoreDistribution(preservationData, moduleOfInterestSize)

Arguments

`preservationData`	the results from PreservationPermutationTest
`moduleOfInterestSize`	the number of genes in your module of interest

Value

A data.frame with Z-summary preservation scores of the module from each permutation and the corresponding module size

Author(s)

Dario Tommasini

Examples

# Remove outlier modules
permutationTestResultsFiltered = lapply(permutationTestResults, function(x) x[!x$is.outlier.module,])

# Find preservation score distribution for a given module size
scores.summary = PreservationScoreDistribution(permutationTestResultsFiltered, 
                                               moduleOfInterestSize = 303)
                                                        
# Remove outlier modules
permutationTestResultsFiltered = lapply(permutationTestResults, function(x) x[!x$is.outlier.module,])

# Find preservation score distribution for a given module size
scores.summary = PreservationScoreDistribution(permutationTestResultsFiltered, 
                                               moduleOfInterestSize = 303)

Run differential module expression

Description

A wrapper to run diffModuleExpression on all the modules in a network

Usage

runDME(
  WGCNAobject,
  design,
  alphaLevel = 0.05,
  testCondition = NULL,
  refCondition = NULL,
  p.adjust = "fdr",
  plot = FALSE,
  test = c("ANOVA", "PERMANOVA"),
  write = FALSE,
  out = NULL
)
runDME(
  WGCNAobject,
  design,
  alphaLevel = 0.05,
  testCondition = NULL,
  refCondition = NULL,
  p.adjust = "fdr",
  plot = FALSE,
  test = c("ANOVA", "PERMANOVA"),
  write = FALSE,
  out = NULL
)

Arguments

`WGCNAobject`	object of class WGCNA with the modules to run DME on
`design`	the sampleTable
`alphaLevel`	level of significance
`testCondition`	the column of the sampleTable to be resolved
`refCondition`	the column of the sampleTable to be used as biological variation
`p.adjust`	adjust for multiple comparisons, argument to pass to p.adjust function
`plot`	generate a plot?
`test`	statistical test to perform, either "ANOVA" or "PERMANOVA"
`write`	write results to a file?
`out`	file name for DME plots, only used if write is TRUE

Value

a data.frame summarizing the results of the analysis

Author(s)

Dario Tommasini

Examples

library(ExperimentHub)
eh = ExperimentHub()
eh_query = query(eh, c("multiWGCNAdata"))
astrocyte_se = eh_query[["EH8223"]]
sampleTable = colData(astrocyte_se)
astrocyte_networks = eh_query[["EH8222"]]
runDME(astrocyte_networks[["combined"]], 
  design = sampleTable,
  p.adjust = "fdr", 
  refCondition = "Region", 
  testCondition = "Disease") 

library(ExperimentHub)
eh = ExperimentHub()
eh_query = query(eh, c("multiWGCNAdata"))
astrocyte_se = eh_query[["EH8223"]]
sampleTable = colData(astrocyte_se)
astrocyte_networks = eh_query[["EH8222"]]
runDME(astrocyte_networks[["combined"]], 
  design = sampleTable,
  p.adjust = "fdr", 
  refCondition = "Region", 
  testCondition = "Disease")

summarizeResults: Summarize results from a results list object

Description

Prints (or writes) a summary of the results from a results list object

Usage

summarizeResults(
  myNetworks,
  results,
  alphaLevel = 0.05,
  write = FALSE,
  outputFile = "results.txt"
)
summarizeResults(
  myNetworks,
  results,
  alphaLevel = 0.05,
  write = FALSE,
  outputFile = "results.txt"
)

Arguments

`myNetworks`	a list of WGCNAobjects
`results`	results list
`alphaLevel`	alpha level of significance
`write`	write to file?
`outputFile`	name of output file, defaults to results.txt

Value

prints a summary of results from the multiWGCNA analysis

TOMFlowPlot

Description

Plots a sankey flow diagram showing the movement of genes from one WGCNA to another WGCNA. Uses the ggalluvial framework.

Usage

TOMFlowPlot(
  WGCNAlist,
  networks,
  toms,
  genes_to_label,
  alpha = 0.1,
  color = "black",
  width = 0.05
)
TOMFlowPlot(
  WGCNAlist,
  networks,
  toms,
  genes_to_label,
  alpha = 0.1,
  color = "black",
  width = 0.05
)

Arguments

`WGCNAlist`	list of WGCNA objects
`networks`	list of network names of length 2
`toms`	a list of TOM distance objects of length 2
`genes_to_label`	genes to label across two networks
`alpha`	alpha of flows
`color`	color of flows
`width`	width of the strata

Value

a ggplot object

Author(s)

Dario Tommasini

topNGenes: Top N genes of a module

Description

Returns the top N number of genes of a module. All genes returned if no number is specified. Genes are in order of intramodular connectivity.

Usage

topNGenes(WGCNAobject, module, nGenes = NULL)
topNGenes(WGCNAobject, module, nGenes = NULL)

Arguments

`WGCNAobject`	an object of class WGCNA
`module`	the name of the module in WGCNAobject
`nGenes`	an integer from 1 to module size; returns all genes if left NULL

Value

a character vector of the genes/probes in the module

Examples

library(ExperimentHub)
eh = ExperimentHub()
eh_query = query(eh, c("multiWGCNAdata"))
astrocyte_networks = eh_query[["EH8222"]]
topNGenes(astrocyte_networks$EAE, "EAE_015", nGenes = 10)

library(ExperimentHub)
eh = ExperimentHub()
eh_query = query(eh, c("multiWGCNAdata"))
astrocyte_networks = eh_query[["EH8222"]]
topNGenes(astrocyte_networks$EAE, "EAE_015", nGenes = 10)

The WGCNA Class

Description

The WGCNA class is the main class used in multiWGCNA to store results from a weighted gene co-expression nework analysis. These include the original unaltered expression data used as input, connectivity metrics, module assignment, input sample conditions, trait

Value

Slots

datExpr: The expression data, connectivity data, and module assignment
conditions: A data.frame with integer conditions for WGCNA
trait: A data.frame showing pearson correlation values to traits
moduleEigengenes: A data.frame of module eigengenes for each module across samples
outlierModules: A vector of modules classified by our algorithm as being driven by sample outliers

Package 'multiWGCNA'

Help Index

Best matching modules

Description

Usage

Arguments

Value

Author(s)

Examples

cleanDatExpr

Description

Usage

Arguments

Value

Author(s)

Examples

Coexpression Line Graph

Description

Usage

Arguments

Value

Author(s)

Examples

computeOverlapsFromWGCNA

Description

Usage

Arguments

Value

Author(s)

Examples

constructNetworks: Construct all the weighted gene correlation networks

Description

Usage

Arguments

Value

Author(s)

Examples

Differential co-expresison analysis

Description

Usage

Arguments

Value

Author(s)

Examples

Differential module expression

Description

Usage

Arguments

Value

Examples

Draw multiWGCNA network

Description

Usage

Arguments

Value

Author(s)

Examples

Get expression data

Description

Usage

Arguments

Value

Author(s)

Examples

getPreservation

Description

Usage

Arguments

Value

Author(s)

iterate: Iterate function across networks

Description

Usage

Arguments

Value

Author(s)

Examples

Generate a trait table from a sample table

Description

Usage