bcdissimilarity: bcdissimilarity

In OxfordCMS/OCMSutility: Utility functions for OCMS

bcdissimilarity

Description

Calculate Bray-Curtis Dissimilarity for samples within and between groups.

Usage

bcdissimilarity(relab_mat, met, sample_id, var)

Arguments

relab_mat	data.frame or matrix. relative abundance table with samples in columns and features in rows; feature IDs as rownames
met	dataframe. metadata table
sample_id	string. Sample identifier (column name in met)
var	string. Metadata variable (column name in met) that specifies groups. set to NULL to bypass between/within group assignment

Value

list. bc_df is long dataframe of samples being compared and corresponding distances. bc_dist matrix of bray curtis distances

Examples

data(dss_example)

count_table <- dss_example$merged_abundance_id %>% column_to_rownames('featureID')
relab_table <- relab(count_table)

met_table <- dss_example$metadata

# bray curtis for one metadata variable
bc_result <- bcdissimilarity(relab_table, met_table, 'sampleID','Phenotype')

pdata <- bc_result$bc_df
p_phen <- ggplot(pdata, aes(x=value, y=dist)) +
   geom_violin() +
   theme_bw(14) +
   ylab('Bray-Curtis Dissimilarity') +
   xlab('Phenotype')

for multiple metadata variables
bc_data <- c()
for(var in c('Phenotype','Genotype')) {
    bc_result <- bcdissimilarity(relab_table, met_table, 'sampleID', var)
    bc_data <- rbind(bc_data, bc_result$bc_df)
}

p_bc <- ggplot(bc_data, aes(x=comparison, y=dist)) +
   geom_violin() +
   theme_bw(14) +
   facet_wrap(~met_var) +
   ylab('Bray-Curtis Dissimilarity') +
   theme(axis.title.x=element_blank())

OxfordCMS/OCMSutility documentation built on Feb. 27, 2025, 8:19 p.m.