Mofify Junction_NanoSplicer.R to take empty files and dataframes

src/.docker_modules/r-bolero/1.0/Junctions_NanoSplicer.R

@@ -2,6 +2,7 @@
 
 ################################################################################
 ### NEED TO ADD A CASE OF NO SPLICED-VARIANTS ARE IDENTIFIED !!!!!!!!!!!!!!! ###
+### PROBLEM IF THERE IS ONLY MULTIPLE JUNCTIONS CONTAINING READS ###############
 ################################################################################
 library(ggplot2, quietly = TRUE)
 library(tidyr, quietly = TRUE)
@@ -25,8 +26,16 @@ colnames(reads_pos) <- c("id", reads_pos[1,2:length(reads_pos[1,])])
 reads_pos <- reads_pos[2:length(reads_pos$id),]
 
 # Load Nanosplicer results:
-df <- read.csv(opt$jwr)
-colnames(df)[1] <- "juncNumber"
+if (file.exists(opt$jwr)) {
+  df <- read.csv(opt$jwr)
+  colnames(df)[1] <- "juncNumber"
+} else { # Define column names
+  column_names <- c("juncNumber", "id", "mapQ", "transcript_strand", "chrID", "loc", "JAQ")
+  
+  # Create an empty dataframe
+  df <- data.frame(matrix(ncol = length(column_names), nrow = 0))
+  colnames(df) <- column_names 
+}
 
 # split donor and acceptor positions:
 df <- df %>%
@@ -164,15 +173,14 @@ SP_assignation_single <- function(site, promoter) {
   }
 }
 
+
 #dplyr version:
-single_junction <- single_junction %>% dplyr::group_by(id) %>%
-  dplyr::mutate(SP_name = SP_assignation_single(junction, promoter))
+single_junction <- single_junction %>% dplyr::group_by(id) 
 
-#plyr version
-# single_junction <- ddply(single_junction,
-#                          .(id), 
-#                          mutate,
-#                          SP_name = SP_assignation_single(junction, promoter))
+if (length(single_junction$promoter) != 0) {
+  single_junction <- single_junction %>% 
+    dplyr::mutate(SP_name = SP_assignation_single(junction, promoter))
+} else { single_junction$SP_name = "" }
 
 SP_assignation_multiple <- function(read_id, combinaison, promoter) {
   if (promoter == "pgRNA" | promoter == "preCore") {
@@ -283,7 +291,9 @@ df_combinaison <- data.frame(matrix(nrow = 0, ncol = 2))
 colnames(df_combinaison) <- c("id", "SP_name")
 
 for (read_id in list_read_multiple) {
-  SP_name_computed <- SP_assignation_multiple(read_id, tmp[tmp$id == read_id,]$junction, tmp[tmp$id == read_id,]$promoter[1])
+  SP_name_computed <- SP_assignation_multiple(read_id, 
+                                              tmp[tmp$id == read_id,]$junction, 
+                                              tmp[tmp$id == read_id,]$promoter[1])
   res_vector <- data.frame(t(c(read_id, SP_name_computed)))
   colnames(res_vector) <- colnames(df_combinaison)
   df_combinaison <- rbind(df_combinaison, res_vector)
@@ -296,7 +306,8 @@ multiple_junction <- merge(multiple_junction, df_combinaison, by="id")
 df_SPvariants <- rbind(as.data.frame(single_junction), multiple_junction)
 
 SP_variant_unique <- df_SPvariants %>% select(id, SP_name)
-SP_variant_unique <- SP_variant_unique[!duplicated(SP_variant_unique$id),] # distinct(SP_variant_unique, id)
+SP_variant_unique <- SP_variant_unique[!duplicated(SP_variant_unique$id),] 
+# distinct(SP_variant_unique, id)
 
 write.table(df_SPvariants, 
             paste0(opt$barcode, "_identified_SPvariants.csv"),