Drop unnecessary dependencies

DESCRIPTION

@@ -15,12 +15,9 @@ Imports:
     GAMBLR.data,
     ggplot2,
     ggthemes,
-    philentropy,
     readr,
-    stringr,
     tibble,
-    tidyr,
-    workflowr
+    tidyr
 Remotes:
     morinlab/GAMBLR.data
 Encoding: UTF-8

NAMESPACE

@@ -19,6 +19,7 @@ export(get_template_wildcards)
 export(get_unmatched_normals)
 export(grob_wildcards)
 export(handle_metadata)
+export(kl_divergence)
 export(maf_header)
 export(normalize_expression_data)
 export(rainfall_conv)
@@ -30,29 +31,10 @@ export(subset_cnstates)
 export(theme_Morons)
 export(trim_scale_expression)
 export(vc_nonSynonymous)
-export(web_initialize_gambl_site)
 import(GAMBLR.data)
 import(dplyr)
 import(ggplot2)
 import(readr)
-import(stringr)
 import(tibble)
 import(tidyr)
-import(workflowr)
-importFrom(dplyr,left_join)
 importFrom(ggthemes,theme_foundation)
-importFrom(philentropy,KL)
-importFrom(stats,end)
-importFrom(stats,quantile)
-importFrom(stats,start)
-importFrom(stringr,str_c)
-importFrom(stringr,str_extract)
-importFrom(stringr,str_remove)
-importFrom(stringr,str_remove_all)
-importFrom(tidyr,unnest_auto)
-importFrom(utils,head)
-importFrom(utils,read.csv)
-importFrom(utils,read.socket)
-importFrom(utils,tail)
-importFrom(utils,write.socket)
-importFrom(utils,write.table)

R/compare_coding_mutation_pattern.R

@@ -8,6 +8,7 @@
 #'
 #' @return list
 #'
+#' @import dplyr
 #' @export
 compare_coding_mutation_pattern = function(maf_df1,maf_df2,gene){
   if(missing(maf_df1) | missing(maf_df2)){
@@ -17,9 +18,9 @@ compare_coding_mutation_pattern = function(maf_df1,maf_df2,gene){
     stop("Must provide the Hugo_Symbol of a single gene that is present in both maf files")
   }
   missense_positions1 = dplyr::filter(maf_df1,Hugo_Symbol==gene,!Variant_Classification %in% c("Silent","Splice_Site","Splice_Region"),Variant_Type=="SNP") %>%
-    pull(HGVSp_Short) %>% str_remove_all("p.\\w") %>% str_extract("\\d+") %>% as.numeric()
+    pull(HGVSp_Short) %>% gsub("p\\.\\w", "", .) %>% regmatches(., regexpr("\\d+", .)) %>% as.numeric()
   missense_positions2 = dplyr::filter(maf_df2,Hugo_Symbol==gene,!Variant_Classification %in% c("Silent","Splice_Site","Splice_Region"),Variant_Type=="SNP") %>%
-    pull(HGVSp_Short) %>% str_remove_all("p.\\w") %>% str_extract("\\d+") %>% as.numeric()
+    pull(HGVSp_Short) %>% gsub("p\\.\\w", "", .) %>% regmatches(., regexpr("\\d+", .)) %>% as.numeric()
  if(length(missense_positions1)==0 | length(missense_positions2)==0 ){
    message(paste("no mutations for",gene,"in one or both data sets"))
    return(list(kl=15))
@@ -35,6 +36,12 @@ compare_coding_mutation_pattern = function(maf_df1,maf_df2,gene){
   all_counts = dplyr::select(full_df,-position) %>% t()
   all_counts[1,]=all_counts[1,]/sum(all_counts[1,])
   all_counts[2,]=all_counts[2,]/sum(all_counts[2,])
-  kl_out = KL(all_counts)
+
+  # Normalize the rows to turn counts into probabilities
+  P <- all_counts[1, ] / sum(all_counts[1, ])
+  Q <- all_counts[2, ] / sum(all_counts[2, ])
+
+  kl_out <- kl_divergence(P, Q)
+
   return(list(df=full_df,kl=unname(kl_out)))
 }

R/gene_mutation_tally.R

@@ -17,6 +17,7 @@
 #'
 #' @return data frame
 #'
+#' @import dplyr
 #' @export
 gene_mutation_tally = function(maf_df,these_samples_metadata,these_genes,grouping_variable="cohort"){
   meta = dplyr::select(these_samples_metadata,sample_id,{{grouping_variable}})

R/get_gambl_colours.R

@@ -18,7 +18,7 @@
 #'
 #' @return A named vector of colour codes for lymphgen classes and pathology.
 #'
-#' @import dplyr stringr tidyr
+#' @import dplyr tidyr
 #' @export
 #'
 #' @examples
@@ -318,7 +318,7 @@ get_gambl_colours = function(classification = "all",
     everything = c(everything, all_colours[[this_group]])
   }
   #return matching value from lowercase version of the argument if it exists
-  lc_class = stringr::str_to_lower(classification)
+  lc_class = tolower(classification)
   if(return_available){
     return(names(all_colours))
   }

R/get_template_wildcards.R

@@ -16,6 +16,6 @@ get_template_wildcards = function(parent_key,
   }else{
     wildcard_string = config::get(paste0(parent_key,"_wildcards"))[template_key]
   }
-  wildcards = stringr::str_split(wildcard_string,",")
+  wildcards = strsplit(wildcard_string,",")
   return(unlist(wildcards))
 }

R/grob_wildcards.R

@@ -8,7 +8,7 @@
 #'
 #' @export
 grob_wildcards = function(wildcarded_string){
-  wildcards = unlist(stringr::str_extract_all(wildcarded_string,"\\{[^\\{]+\\}"))
-  wildcards = stringr::str_remove_all(wildcards,"\\{") %>%  stringr::str_remove_all(.,"\\}")
+  wildcards = unlist(regmatches(wildcarded_string, gregexpr("\\{[^\\{]+\\}", wildcarded_string)))
+  wildcards = gsub("\\{", "", wildcards) %>% gsub("\\}", "", .)
   return(wildcards)
 }

R/kl_divergence.R

@@ -0,0 +1,36 @@
+#' Calculate Kullback-Leibler Divergence
+#'
+#' This function computes the Kullback-Leibler (KL) divergence between two
+#'      probability distributions, with an optional small constant (epsilon)
+#'      added to avoid zero probabilities, which would otherwise cause division
+#'      by zero or undefined logarithms.
+#'
+#' @param P A numeric vector representing the first probability distribution.
+#'      The sum of "P" should be 1, but the function will normalize it if
+#'      necessary.
+#' @param Q A numeric vector representing the second probability distribution.
+#'      The sum of "Q" should be 1, but the function will normalize it if
+#'      necessary.
+#' @param epsilon A small positive number (default = 1e-7) to be added to each
+#'      probability in P and Q to avoid zero probabilities. This helps to
+#'      prevent division by zero or log(0).
+#'
+#' @return float
+#'
+#' @examples
+#' P <- c(0.1, 0.4, 0.3, 0.2)
+#' Q <- c(0.2, 0.3, 0.4, 0.1)
+#'
+#' kl_divergence(P, Q)
+#'
+#' @export
+kl_divergence <- function(P, Q, epsilon = 1e-7) {
+    P <- P + epsilon
+    Q <- Q + epsilon
+
+    P <- P / sum(P)
+    Q <- Q / sum(Q)
+
+    # KL divergence formula: sum(P * log(P / Q))
+    return(sum(P * log(P / Q), na.rm = TRUE))
+}

R/sanity_check_metadata.R

@@ -6,7 +6,7 @@
 #'
 #' @return A table.
 #'
-#' @import tibble readr dplyr
+#' @import tibble readr dplyr tidyr
 #'
 #'
 #' @examples