merger.py

#!/usr/bin/env python
# -*- coding: utf-8 -*-
from itertools import *
from operator import itemgetter
from fnmatch import fnmatch
import sys, csv

def group_wc_in_full(d_wc, merge_full):
	reslist = []
	res_no_matches = []
	for wc in d_wc:
		matches = 0
		for full in merge_full:
			if match(full[1], wc[1]):
				matches += 1
		if matches == 1:
			cont = 0
			for full in merge_full:
				if match(full[1], wc[1]):
					new_key = full[0] + ", (" + wc[0] + ")"
					merge_full[cont] = (new_key,full[1])
				cont += 1
		if matches > 1:
			cont = 0
			for full in merge_full:
				if match(full[1], wc[1]):
					new_key = full[0] + ", [" + wc[0] + "]"
					merge_full[cont] = (new_key,full[1])
				cont += 1
		if matches == 0:
			res_no_matches.append(wc)
	return merge_full, res_no_matches

def group_no_wc(mergelist):
	reslist = []
	total_keys = []
	for i in range(len(mergelist)):
		isMatch = False
		compared = False
		keys = []
		anyval = None
		keys.append(mergelist[i][0])
		if mergelist[i][0] in total_keys:
			continue
		for j in range(i + 1, len(mergelist) ):
			if mergelist[j][0] in total_keys:
				continue
			compared = True
			if mergelist[i][1] == mergelist[j][1]:
				isMatch = True
				keys.append(mergelist[j][0])
				total_keys.append(mergelist[j][0])
				anyval = mergelist[j][1]
		if compared:
			if isMatch:
				str_keys = ", ".join(keys)
				new_item = (str_keys, anyval)
				reslist.append(new_item)
			else:
				reslist.append(mergelist[i])
	return reslist

def match(v1, v2):
	if not "-" in v1 and not "-" in v2:
		return v1 == v2
	v1_reg = v1.replace("-","?")
	cont = 0
	for s in v2:
		if s == "-":
			v1_reg = v1_reg[:cont] + "?" + v1_reg[cont+1:]
		cont += 1
	return fnmatch(v2, v1_reg)

def group_nonmatched(nonmatched):
	pivot = nonmatched[0]
	res = []
	res.append(pivot)
	for i in range(1, len(nonmatched)):
		count = 0
		keys = []
		keys.append(nonmatched[i][0])
		for j in range(len(res)):
			try:
			    a = res[j]
			except IndexError:
			    continue
			if match(nonmatched[i][1], res[j][1]):
				count += 1
				keys.append(str(res[j][0]))
				del(res[j])
		if count == 0:
			res.append(nonmatched[i])
		else:
			new_tuple = (", ".join(keys), nonmatched[i][1] )
			res.append(new_tuple)
	for i in range(len(res)):
		res[i] = ("{" + res[i][0] + "}", res[i][1])
	return res

d = {	'k1': 'AABA',
 		'k2': 'AABA',
		'k3': 'ABAA',
		'k4': 'AABA',
		'k5': 'A--A',
		'k6': 'AB-A',
		'k7': 'BB-A',# no matched -->
		'k8': 'B--A',
		'k9': 'B-BA',
		'k10':'B--B',
		'k11':'BA-B',
		'k12':'AAA-',
		'k13':'AA-B',
		'k14':'AB-B',
		'k14':'-B-B',
		}

input_file_name = sys.argv[1]
output_file_name = sys.argv[2]

input_file = list(csv.reader(open(input_file_name, 'rb'), delimiter=','))


d = {}
for line in input_file:
	str_line = "".join(line[3:len(line)])
	d[line[0]] = str_line



d_full = list((key, value) for key, value in d.iteritems())
d_no_wc = list((key,value) for key, value in d.iteritems() if not "-" in value)
d_wc = list((key,value) for key, value in d.iteritems() if "-" in value)



#1- merge_full <= MERGE d_no_wc
merge_full = group_no_wc(d_no_wc)

#2- merge_full_wc <= MERGE d_wc in merge_full
merge_full_wc, wc_no_match = group_wc_in_full(d_wc, merge_full)

#3 merge nonmatched
if len(wc_no_match) > 0:
	merged_wc_no_match = group_nonmatched(wc_no_match)
else:
	merged_wc_no_match = []

final = merged_wc_no_match + merge_full_wc

idx = 0
for row in final:
	positions = {}
	ids = row[0]
	ids = ids.replace("(","").replace(")","")
	ids = ids.replace("[","").replace("]","")
	ids = ids.replace("{","").replace("}","")
	ids = ids.replace(" ","")
	ids = ids.split(',')
	for id in ids:
		id_clean = id.replace(" ","")
		for line in input_file:
			if line[0] == id_clean:
				positions.setdefault(line[1], []).append( (id_clean,line[2]) )
	#get most common element in list (dict keys)
	chromosome = max(set(positions.keys()), key=positions.keys().count)

	_id, position = min(positions[chromosome], key= lambda x: x[1])
	final[idx] += (chromosome, position, _id, )
	idx += 1

output_file = open(output_file_name, 'w')
for item in final:
	row = []
	row.append(item[0]) # ids
	row.append(item[2]) #chromosome
	row.append(item[3]) #position
	row.append(item[4]) #id from position and chr
	for c in item[1]:
		row.append(c)
	output_file.write("\t".join(row) + "\n")