plot_cosine_heatmap: Plot cosine similarity heatmap

In UMCUGenetics/MutationalPatterns: Comprehensive genome-wide analysis of mutational processes

Plot cosine similarity heatmap

Description

Plot pairwise cosine similarities in a heatmap.

Usage

plot_cosine_heatmap(
  cos_sim_matrix,
  col_order = NA,
  row_order = NA,
  cluster_rows = TRUE,
  cluster_cols = FALSE,
  method = "complete",
  plot_values = FALSE
)

Arguments

cos_sim_matrix	Matrix with pairwise cosine similarities. Result from cos_sim_matrix
col_order	Character vector with the desired order of the columns names for plotting. Optional.
row_order	Character vector with the desired order of the row names for plotting. Optional.
cluster_rows	Hierarchically cluster rows based on euclidean distance. Default = TRUE.
cluster_cols	Hierarchically cluster cols based on euclidean distance. Default = FALSE.
method	The agglomeration method to be used for hierarchical clustering. This should be one of "ward.D", "ward.D2", "single", "complete", "average" (= UPGMA), "mcquitty" (= WPGMA), "median" (= WPGMC) or "centroid" (= UPGMC). Default = "complete".
plot_values	Plot cosine similarity values in heatmap. Default = FALSE.

Value

Heatmap with cosine similarities

See Also

mut_matrix, cos_sim_matrix

Examples

## See the 'mut_matrix()' example for how we obtained the mutation matrix:
mut_mat <- readRDS(system.file("states/mut_mat_data.rds",
  package = "MutationalPatterns"
))

## Get signatures
signatures <- get_known_signatures()

## Calculate the cosine similarity between each signature and each 96 mutational profile
cos_matrix <- cos_sim_matrix(mut_mat, signatures)


## Plot the cosine similarity between each signature and each sample with hierarchical
## clustering of samples and signatures.
plot_cosine_heatmap(cos_matrix, cluster_rows = TRUE, cluster_cols = TRUE)

## In the above example, clustering is performed on the similarities of the samples with
## the signatures. It's also possible to cluster the signatures and samples on their (96) profile.
## This will generally give better results
## If you use the same signatures for different analyses,
## then their order will also be consistent.
hclust_cosmic <- cluster_signatures(signatures, method = "average")
cosmic_order <- colnames(signatures)[hclust_cosmic$order]
hclust_samples <- cluster_signatures(mut_mat, method = "average")
sample_order <- colnames(mut_mat)[hclust_samples$order]
## Plot the cosine heatmap using this given signature order.
plot_cosine_heatmap(cos_matrix,
  cluster_rows = FALSE, cluster_cols = FALSE,
  row_order = sample_order, col_order = cosmic_order
)

## You can also plot the similarity of samples with eachother
cos_matrix <- cos_sim_matrix(mut_mat, mut_mat)
plot_cosine_heatmap(cos_matrix, cluster_rows = TRUE, cluster_cols = TRUE)


## It's also possible to add the actual values in the heatmap.
plot_cosine_heatmap(cos_matrix, cluster_rows = TRUE, cluster_cols = TRUE, plot_values = TRUE)

UMCUGenetics/MutationalPatterns documentation built on Nov. 24, 2022, 4:31 a.m.