fixing the bug with footprint length

bin/1.1_footprint_extraction/footprints_extraction.py

@@ -154,6 +154,7 @@ def save_footprint(footprint_count, footprint_scores, peak_footprints, chromosom
 				old_start = peak_footprints[existing_footprint_name]['start']
 				old_end = peak_footprints[existing_footprint_name]['end']
 				old_score = peak_footprints[existing_footprint_name]['score']
+				old_length = peak_footprints[existing_footprint_name]['len']
 
 				if footprint_start >= old_start and footprint_start <= old_end: #the start of the new footprint is between the start and end of an old footprint
 					if footprint_end > old_end: #the new footprint is not completely inside the old one
@@ -163,19 +164,21 @@ def save_footprint(footprint_count, footprint_scores, peak_footprints, chromosom
 
 						peak_footprints[existing_footprint_name]['end'] = footprint_end
 						peak_footprints[existing_footprint_name]['score'] = footprint_score
+						peak_footprints[existing_footprint_name]['len'] = footprint_end - old_start
 						#we can not update the max_pos as we do not have the information about scores array of the existing footprint
 					#else: the new footprint is completely inside the old one, do nothing
 					save_current_footprint = False
 					break
 
 				elif footprint_end >= old_start and footprint_end <= old_end: #the end of the new footprint is between the start and end of an old footprint
-					if footprint_start < old_start: #the new footprint is not completely inside the old one
+					if footprint_start < old_start: #the new footprint is not completely inside the old one						
 						#update the information about the existing footprint
 						#find the average of both scores
 						footprint_score = (peak_footprints[existing_footprint_name]['score'] + footprint_score) / 2
 
 						peak_footprints[existing_footprint_name]['start'] = footprint_start
 						peak_footprints[existing_footprint_name]['score'] = footprint_score
+						peak_footprints[existing_footprint_name]['len'] = old_end - footprint_start
 					#else do nothing
 					save_current_footprint = False
 					break
@@ -270,6 +273,11 @@ def search_in_window(peak_footprints, footprint_count, chromosom, peak_start, pe
 
 	return peak_footprints, footprint_count
 
+def check_for_overlap_and_merge(peak_footprints):
+	for footprint_to_check in peak_footprints.keys():
+		print(footprint_to_check, peak_footprints[footprint_to_check])
+	return peak_footprints
+
 #this function uses the information provided from the .bed file to look for footprints within the peaks of interest
 #as input the information from the original bed file, as well as bigwig file is needed
 #the optional parameters window_length, step and percentage are needed as well to use the sliding window algorithm and work with the "background" score
@@ -295,9 +303,12 @@ def find_peaks_from_bw(bed_dictionary, bw_file, window_length, step, percentage)
 			peak_end = int(positions[1])
 
 			scores_in_peak = np.nan_to_num(np.array(list(bw_open.values(chromosom, peak_start, peak_end)))) #save the scores to an array
-
+			print()
 			peak_footprints, footprint_count = search_in_window(peak_footprints, footprint_count, chromosom, peak_start, peak_end, scores_in_peak, window_length, bed_dictionary[header], step, percentage)
 
+			#double check for overlaps and possibly merging of footprints having up to 5 bp in between
+			peak_footprints = check_for_overlap_and_merge(peak_footprints)
+
 			for footprint_name in peak_footprints.keys():
 				all_footprints[footprint_name] = all_footprints.get(footprint_name, {})
 				all_footprints[footprint_name] = peak_footprints[footprint_name]