plot_contribution: Plot signature contribution barplot

In MutationalPatterns: Comprehensive genome-wide analysis of mutational processes

Description

Plot contribution of signatures. Can be used on both the results of a NMF and on the results of signature refitting.

Usage

plot_contribution(
  contribution,
  signatures = NA,
  index = NA,
  coord_flip = FALSE,
  mode = c("relative", "absolute"),
  palette = NA
)

Arguments

contribution	Signature contribution matrix
signatures	Signature matrix. Necessary when plotting NMF results in "absolute" mode. It's not necessary in relative mode or when visualizing signature refitting results
index	optional sample subset parameter
coord_flip	Flip X and Y coordinates, default = FALSE
mode	"relative" or "absolute"; to plot the relative contribution or absolute number of mutations, default = "relative"
palette	A color palette like c("#FF0000", "#00FF00", "9999CC") that will be used as colors in the plot. By default, ggplot2's colors are used to generate a palette.

Value

Stacked barplot with contribution of each signature for each sample

See Also

extract_signatures, mut_matrix

Examples

## Extracting signatures can be computationally intensive, so
## we use pre-computed data generated with the following command:
# nmf_res <- extract_signatures(mut_mat, rank = 2)

nmf_res <- readRDS(system.file("states/nmf_res_data.rds",
  package = "MutationalPatterns"
))

## Optionally set column and row names.
colnames(nmf_res$signatures) <- c("Signature A", "Signature B")
rownames(nmf_res$contribution) <- c("Signature A", "Signature B")

## Plot the relative contribution
plot_contribution(nmf_res$contribution)

## Plot the absolute contribution.
## When plotting absolute NMF results, the signatures need to be included.
plot_contribution(nmf_res$contribution,
  nmf_res$signature,
  mode = "absolute"
)


## Only plot a subset of samples
plot_contribution(nmf_res$contribution,
  nmf_res$signature,
  mode = "absolute",
  index = c(1, 2)
)
## Flip the coordinates
plot_contribution(nmf_res$contribution,
  nmf_res$signature,
  mode = "absolute",
  coord_flip = TRUE
)

## You can also use the results of signature refitting.
## Here we load some data as an example
fit_res <- readRDS(system.file("states/snv_refit.rds",
  package = "MutationalPatterns"
))
plot_contribution(fit_res$contribution)

## Or again in absolute mode
plot_contribution(fit_res$contribution,
  mode = "absolute"
)

MutationalPatterns documentation built on Nov. 14, 2020, 2:03 a.m.