####Libraries####
packages = c("ggplot2","tidyr","jsonlite","tidyverse","reshape2","extrafont")
invisible(lapply(packages, library, character.only = TRUE))
options(bitmapType="cairo")

####users selection####
lim_inf = 1 #select the minimum isenrtion size
lim_sup = 700 #select the maximum insertion size

####define json files path####
json_paths <- read.delim("/users/project/precisiontox_008279/grape-MGI_comaprison/insertion_distribution/all_json_pipeline.db",header = FALSE,sep = "\t")
json_paths <- json_paths[,c(1,3)]

for(j in 1:nrow(json_paths))
{
print(paste("processing sample",j,"of",nrow(json_paths),sep = " "))
json_sample <- json_paths[j,2]
####read files with jsonlite and retrieve insert size#### 
sample <- fromJSON(json_sample)
sample_insertion_size <-sample$general$pairs$insert_sizes
####trasform insert size in data.frame####
insert_size_df <- data.frame(matrix(nrow = 50000, ncol = 1))
colnames(insert_size_df) <- c("insertion length")
n=1
for(i in 1:nrow(insert_size_df))
   {
   if(i == as.numeric(names(sample_insertion_size[n])))
    {
     insert_size_df[i,1] <- sample_insertion_size[n]
     n=n+1
   }
}
####distribution plot####
insert_size_df[is.na(insert_size_df)] <- 0
insert_size_df <- melt(insert_size_df)
insert_size_df <- insert_size_df[lim_inf:lim_sup,]
insert_size_df$size <- as.numeric(rownames(insert_size_df))
#plot
distribution_plot <- ggplot(insert_size_df,aes(x=size,y=value,col=variable)) + geom_bar(stat='identity',show.legend = FALSE) +
xlab("insertion length") + xlim(lim_inf,lim_sup) + ylab("count") + ggtitle(paste(json_paths[j,1]," insertion length distribution",sep = " "))
####save png####
ggsave(filename = paste("/users/project/precisiontox_008279/grape-MGI_comaprison/insertion_distribution/results/section_0-700/",json_paths[j,1],"_insertion_plot.png",sep = ""),
       plot = distribution_plot)
}