max_pos calculation is done in one Rscript

bin/1.2_filter_motifs/compareBed_runinfo.R

@@ -27,12 +27,29 @@ output <- args[5]
 output_stats <- args[6]
 
 data_filtered <- fread(input_filtered, sep='\t')
+# data is the initial data before any comparisons have been done (-d parameter of compareBed.sh)
+data <- fread(input_raw, sep='\t')
 
 # check if data has less than 9 columns
 if (ncol(data_filtered) < 9) {
 	stop("footprint file has less than 9 columns. exiting.")
 }
 
+# define needed col names
+names(data)[2] <- "old_start"
+names(data_filtered)[c(2, 3, 8)] <- c("start", "end", "max_pos")
+# save corect column order
+correct_order <- c(names(data_filtered), names(data)[2])
+# compute new maxpos (relevant for splitted footprints)
+data_filtered <- merge(x=data_filtered, y=data[,c(2,4)], by.x=names(data_filtered)[4], by.y=names(data)[4], all.x=TRUE, all.y=FALSE, sort=FALSE)
+data_filtered <- data_filtered[, correct_order, with = FALSE]
+data_filtered[, max_pos := max_pos + old_start - start][, old_start := NULL]
+
+# adding column "contains_maxpos", containing flag (0 or 1)
+# max_pos is the position of maximum score of a footprint
+data_filtered <- cbind(data_filtered, contains_maxpos = 0)
+data_filtered[max_pos >= 0 & start + max_pos <= end, contains_maxpos := 1]
+
 # remove sequences that are smaller than minimum (parameter)
 # remove sequences that are longer than maximum (parameter)
 data_filtered <- data_filtered[which(data_filtered[[3]] - data_filtered[[2]] >= min),]
@@ -50,17 +67,8 @@ data_filtered[match(duplicants, names), name := paste0(name,".0")]
 # recalculate length of sequences
 data_filtered[[7]] <- data_filtered[[3]] - data_filtered[[2]]
 
-# adding column "contains_maxpos", containing flag (0 or 1)
-# max_pos is the position of maximum score of a footprint
-data_filtered <- cbind(data_filtered, contains_maxpos = 0)
-data_filtered$contains_maxpos[intersect(which(data_filtered[[2]] <= data_filtered[[8]]), which(data_filtered[[3]] > data_filtered[[8]]))] = 1
-data_filtered[[8]] <- data_filtered[[8]] - data_filtered[[2]]
-
 fwrite(data_filtered, output, col.names=FALSE, quote = FALSE, sep = '\t')
 
-# data is the initial data before any comparisons have been done (-d parameter of compareBed.sh)
-data <- fread(input_raw, sep='\t')
-
 # some statistics about the bedtool comparisons are stored in FilterMotifs.stats
 # number of nucleotides input
 sum_data <- sum(data[[3]]-data[[2]])