MOFA2 PLOTTING

Explore a model of a multi-omic data set with some basics plotting.

URL: https://www.bioconductor.org/packages/release/bioc/html/MOFA2.html

Example

This wrapper can be used in the following way:

rule test_all:
    input:
        "results/data_overview.png",
        "results/data_variance_explained.png",
        "results/data_feature_weights.png",
        "results/data_data_heatmap.png",


rule mofa2_plotting:
    input:
        "{data}.hdf5",
    output:
        overview="results/{data}_overview.png",  # no requirements
        variance_explained="results/{data}_variance_explained.png",  # no requirements
        feature_weights="results/{data}_feature_weights.png",  # requires params: `view`, `factor`, `nfeatures`
        data_heatmap="results/{data}_data_heatmap.png",  # requires params: `view`, `factor`, `features`
    log:
        "log/{data}.log",
    params:
        view="view_0",  # the name of the view to be plotted
        factor=1,  # which factor to plot
        features=20,  # the number of features to plot
        nfeatures=10,  # the number of features to highlight
    wrapper:
        "v9.4.0/bio/mofa2/plotting"

Note that input, output and log file paths can be chosen freely.

When running with

snakemake --use-conda

the software dependencies will be automatically deployed into an isolated environment before execution.

Notes

Each of the plots will be output as a separate PNG file. Choose the plots by setting the output accordingly.
Make sure to set the params required for the desired output.

Software dependencies

bioconductor-mofa2=1.20.2

Input/Output

Input:

An HDF5-file with the trained model.

Output:

overview: Optional. PNG file. Shows the number of views (rows) and groups (columns), what their corresponding dimensionalities are, and how much missing data they have. Doesn’t require any params to be set.
variance_explained: Optional. PNG file. Shows how much of the variance is explained by each factor in each data modality. Includes all factors. Doesn’t require any params to be set.
feature_weights: Optional. PNG file. The weights provide a score for how strong each feature relates to each factor. Features with no association with the factor have values close to zero, while features with strong association with the factor have large absolute values. The sign of the weight indicates the direction of the effect: a positive weight indicates that the feature has higher levels in the cells with positive factor values, and vice versa. Requires the following params to be set: view, factor, nfeatures.
data_heatmap: Optional. PNG file. Heatmap of observations to allow for an observation of the coordinated heterogeneity that MOFA captures in the original data. Top features are selected by their weight in the selected factor. Samples are ordered according to their corresponding factor value. Requires the following params to be set: view, factor, features.

Params

view: String that specifies the name of the view from the data set that will be plotted.
factor: Integer that specifies the index of the factor that will be plotted. Required for feature_weights, data_heatmap.
features: Integer that specifies the number of features to be plotted. Required for data_heatmap plot.
nfeatures: Integer that specifies the number of features to be highlighted. Required for feature_weights plot.

Authors

Simon Sack

Code

#!/bin/R

# load libraries
library(MOFA2)
library(ggplot2)

# if log file is provided, write log to that file
if (length(snakemake@log) > 0) {
  log <- file(snakemake@log[[1]], open = "wt")
  sink(log)
  sink(log, type = "message")
}


# cast input and output path as character to avoid errors
input_path <- as.character(snakemake@input[[1]])

# load a MOFA2 model from an hdf5 file
model <- load_model(input_path)


# setting the variables customisable with params

if ("view" %in% names(snakemake@params)) {
  view <- snakemake@params[["view"]]
} else {
  view <- "view_0"
}

if ("factor" %in% names(snakemake@params)) {
  factor <- snakemake@params[["factor"]]
} else {
  factor <- 1
}

if ("features" %in% names(snakemake@params)) {
  features <- snakemake@params[["features"]]
} else {
  features <- 10
}

if ("nfeatures" %in% names(snakemake@params)) {
  nfeatures <- snakemake@params[["nfeatures"]]
} else {
  nfeatures <- 10
}

# creating the requested plots

# overview plot
if ("overview" %in% names(snakemake@output)) {
  overview_path <- as.character(snakemake@output[["overview"]])
  p <- plot_data_overview(model)

  # write plot to file
  ggsave(plot = p, filename = overview_path)
}

# variance_explained plot
if ("variance_explained" %in% names(snakemake@output)) {
  variance_explained_path <- as.character(snakemake@output[["variance_explained"]])
  p <- plot_variance_explained(model, x="view", y="factor")

  # write plot to file
  ggsave(plot = p, filename = variance_explained_path)
}

# feature weights plot
if ("feature_weights" %in% names(snakemake@output)) {
  feature_weights_path <- as.character(snakemake@output[["feature_weights"]])
  p <- plot_weights(model,
    view = view, # which view to plot
    factor = factor, # which factor to plot
    nfeatures = nfeatures, # number of features to highlight
    scale = TRUE, # scale weights from -1 to 1
    abs = FALSE # take the absolute value?
  )

  # write plot to file
  ggsave(plot = p, filename = feature_weights_path)
}

# covariation patterns

## covariation patterns heatmap
if ("data_heatmap" %in% names(snakemake@output)) {
  data_heatmap_path <- as.character(snakemake@output[["data_heatmap"]])
  p <- plot_data_heatmap(model,
    view = view, # which view to plot
    factor = factor, # which factor to plot
    features = features, # how many features to plot

    cluster_rows = TRUE, cluster_cols = FALSE,
    show_rownames = TRUE, show_colnames = FALSE
  )

  # write plot to file
  ggsave(plot = p, filename = data_heatmap_path)
}

# checks whether Rplots was created and removes it
if (file.exists("Rplots.pdf")) {
  file.remove("Rplots.pdf")
}