cos_sim_matrix {MutationalPatterns} | R Documentation |
Computes all pairwise cosine similarities between the mutational profiles provided in the two mutation count matrices. The cosine similarity is a value between 0 (distinct) and 1 (identical) and indicates how much two vectors are alike.
cos_sim_matrix(mut_matrix1, mut_matrix2)
mut_matrix1 |
96 mutation count matrix (dimensions: 96 mutations X n samples) |
mut_matrix2 |
96 mutation count matrix (dimensions: 96 mutations X m samples) |
Matrix with pairwise cosine similarities (dimensions: n mutational profiles X m mutational profiles)
mut_matrix
,
fit_to_signatures
,
plot_cosine_heatmap
## You can download mutational signatures from the COSMIC database: # sp_url = http://cancer.sanger.ac.uk/cancergenome/assets/signatures_probabilities.txt # cancer_signatures = read.table(sp_url, sep = "\t", header = T) ## We copied the file into our package for your convenience. filename <- system.file("extdata/signatures_probabilities.txt", package="MutationalPatterns") cancer_signatures <- read.table(filename, sep = "\t", header = TRUE) ## See the 'mut_matrix()' example for how we obtained the mutation matrix: mut_mat <- readRDS(system.file("states/mut_mat_data.rds", package="MutationalPatterns")) ## Match the order to MutationalPatterns standard of mutation matrix order = match(row.names(mut_mat), cancer_signatures$Somatic.Mutation.Type) ## Reorder cancer signatures dataframe cancer_signatures = cancer_signatures[order,] ## Use trinucletiode changes names as row.names ## row.names(cancer_signatures) = cancer_signatures$Somatic.Mutation.Type ## Keep only 96 contributions of the signatures in matrix cancer_signatures = as.matrix(cancer_signatures[,4:33]) ## Rename signatures to number only colnames(cancer_signatures) = as.character(1:30) ## Calculate the cosine similarity between each COSMIC signature and each 96 mutational profile cos_sim_matrix(mut_mat, cancer_signatures)