MOFA (version 1.3.1)

plotDataHeatmap: Plot heatmap of relevant features

Description

Function to plot a heatmap of the input data for relevant features, usually the ones with highest loadings in a given factor.

Usage

plotDataHeatmap(object, view, factor, features = 50,
  includeWeights = FALSE, transpose = FALSE, imputed = FALSE, ...)

Arguments

object

a MOFAmodel object.

view

character vector with the view name, or numeric vector with the index of the view.

factor

character vector with the factor name, or numeric vector with the index of the factor.

features

if an integer, the total number of top features to plot, based on the absolute value of the loading. If a character vector, a set of manually-defined features. Default is 50.

includeWeights

logical indicating whether to include the weight of each feature as an extra annotation in the heatmap. Default is FALSE.

transpose

logical indicating whether to transpose the output heatmap. Default corresponds to features as rows and samples as columns.

imputed

logical indicating whether to plot the imputed data instead of the original data. Default is FALSE.

...

further arguments that can be passed to pheatmap

Value

plots a heatmap of the data for the top features for a given factor and views

Details

One of the first steps for the annotation of a given factor is to visualise the corresponding loadings, using for example plotWeights or plotTopWeights. These functions display the top features that are driving the heterogeneity captured by a factor. However, one might also be interested in visualising the coordinated heterogeneity in the input data, rather than looking at "abstract" weights. This function extracts the top features for a given factor and view, and generates a heatmap with dimensions (samples,features). This should reveal the underlying heterogeneity that is captured by the latent factor. A similar function for doing scatterplots rather than heatmaps is plotDataScatter.

Examples

# NOT RUN {
# Load CLL data
filepath <- system.file("extdata", "CLL_model.hdf5", package = "MOFAdata")
MOFA_CLL <- loadModel(filepath)
# plot top 30 features on Factor 1 in the mRNA view
plotDataHeatmap(MOFA_CLL, view="mRNA", factor=1, features=30)
# without column names (extra arguments passed to pheatmap)
plotDataHeatmap(MOFA_CLL, view="mRNA", factor=1, features=30, show_colnames = FALSE)
# transpose the heatmap
plotDataHeatmap(MOFA_CLL, view="mRNA", factor=1, features=30, transpose=TRUE)
# do not cluster rows (extra arguments passed to pheatmap)
plotDataHeatmap(MOFA_CLL, view="mRNA", factor=1, features=30, cluster_rows=FALSE)
# }