R/CALC-calc_binary.R

In mirnavazquez/RbiMs: R Tools for Reconstruncting Bin Metabolisms

#' @title Calculate presence/absence.
#' @description  Calculate the presence/absence of data in certain pathway.
#' @usage calc_binary (tibble_ko, y_axis, data_experiment = NULL, binary = TRUE,
#' metabolism = FALSE)
#' @param tibble_ko a tibble object from mapping_ko.
#' @param y_axis a character, indicating the pathway to analyze.
#' @param data_experiment optional. a data frame object 
#' containing metadata information.
#' @param binary optional. a logical value, indicating whether you want to
#' calculate presence/absence or abundance of KOs per pathway.
#' @param metabolism optional. a logical value, if TRUE it will show 
#' the metabolism table.
#' @details This function calculates the presence or absence of a pathway by 
#' counting the number of times it appears (or not) in the dataset.
#' @import  tibble dplyr stringr tidyr  rlang
#' @examples
#' # calc_binary (ko_bin_mapp, Pathway, binary = T, metabolism = T)
#' @return A data frame containing the presence/absence (1/0) of each pathway
#' within each bin_name. 
#' calc_binary(ko_bin_mapp, Pathway)    
#' @export

# Sets the syntax of the function------------------------------------------####
calc_binary <- function(tibble_ko,
                        y_axis,
                        data_experiment = NULL,
                        binary = TRUE,
                        metabolism = FALSE){
# Enquoting ---------------------------------------------------------------####
  y_axis_enquo <- enquo(y_axis)
  y_axis_label <- as_label(y_axis_enquo)
# Select data -------------------------------------------------------------####
  data_to_select <- c("Module", "Module_description", "Pathway", 
                      "Pathway_description", "Genes", 
                      "Gene_description", "Enzyme", "Cycle", "Pathway_cycle",
                      "Detail_cycle", "rbims_pathway", "rbims_sub_pathway",
                      "KO", "dbCAN", "domain_name", "Pfam", "PFAM", "INTERPRO")
# Transform from wide to long ---------------------------------------------####
  Kegg_long<- tibble_ko %>%
    pivot_longer(cols = -any_of(data_to_select), 
                 values_to = "Abundance",
                 names_to = "Bin_name") %>%
    distinct()
# Calc binary -------------------------------------------------------------####
  if(binary == TRUE){
    Table_binary <- suppressMessages(Kegg_long %>%
                                       rename(tmp = .data$Abundance)  %>%
                                       select({{y_axis_enquo}}, .data$Bin_name, 
                                              .data$tmp) %>%
                                       distinct() %>%
                                       group_by({{y_axis_enquo}}, 
                                                .data$Bin_name)  %>%
                                       summarise(Presence_absence = 
                                                 sum(.data$tmp)) %>% 
                                       ungroup() %>%
                                       drop_na() %>%
                                       mutate(Presence_absence = 
                                              case_when(.data$Presence_absence 
                                              != "0" ~ "1", TRUE ~ as.character
                                              (.data$Presence_absence))) %>%
                                       mutate_at('Presence_absence', as.integer)) 
  } else if (binary == FALSE) {
    Table_binary <- suppressMessages(Kegg_long %>%
                                       select({{y_axis_enquo}}, .data$Bin_name, 
                                              .data$Abundance) %>%
                                       distinct() %>%
                                       group_by({{y_axis_enquo}}, 
                                                .data$Bin_name) %>%
                                       summarise(Abundance = 
                                                 sum(.data$Abundance)) %>% 
                                       ungroup() %>%
                                       drop_na())
  }
# Join data experiment ----------------------------------------------------####
  if(is.null(data_experiment) == F){
    Table_binary <- Kegg_long %>% left_join(Table_binary, by = "Bin_name")
  }
# Full table ------------------------------------------------------------- ####
  if(metabolism == T){
    Kegg_long <- select(Kegg_long, -.data$Abundance)
    Table_binary <- Table_binary %>%
    left_join(Kegg_long, by = c(y_axis_label, "Bin_name"))
  }
  return(Table_binary)
}