annotate shm_csr.py @ 26:ce25fb581ca3 draft

Uploaded

1 from __future__ import division
2 from collections import defaultdict
3 import re
4 import argparse
5 
6 parser = argparse.ArgumentParser()
7 parser.add_argument("--input",
8 					help="The '7_V-REGION-mutation-and-AA-change-table' and '10_V-REGION-mutation-hotspots' merged together, with an added 'best_match' annotation")
9 parser.add_argument("--genes", help="The genes available in the 'best_match' column")
10 parser.add_argument("--empty_region_filter", help="Where does the sequence start?", choices=['leader', 'FR1', 'CDR1', 'FR2'])
11 parser.add_argument("--output", help="Output file")
12 
13 args = parser.parse_args()
14 
15 infile = args.input
16 genes = str(args.genes).split(",")
17 empty_region_filter = args.empty_region_filter
18 outfile = args.output
19 
20 genedic = dict()
21 
22 mutationdic = dict()
23 mutationMatcher = re.compile("^(.)(\d+).(.),?(.)?(\d+)?.?(.)?(.?.?.?.?.?)?")
24 NAMatchResult = (None, None, None, None, None, None, '')
25 linecount = 0
26 
27 IDIndex = 0
28 best_matchIndex = 0
29 fr1Index = 0
30 cdr1Index = 0
31 fr2Index = 0
32 cdr2Index = 0
33 fr3Index = 0
34 first = True
35 IDlist = []
36 mutationList = []
37 mutationListByID = {}
38 cdr1LengthDic = {}
39 cdr2LengthDic = {}
40 
41 with open(infile, 'r') as i:
42 	for line in i:
43 		if first:
44 			linesplt = line.split("\t")
45 			IDIndex = linesplt.index("Sequence.ID")
46 			best_matchIndex = linesplt.index("best_match")
47 			fr1Index = linesplt.index("FR1.IMGT")
48 			cdr1Index = linesplt.index("CDR1.IMGT")
49 			fr2Index = linesplt.index("FR2.IMGT")
50 			cdr2Index = linesplt.index("CDR2.IMGT")
51 			fr3Index = linesplt.index("FR3.IMGT")
52 			cdr1LengthIndex = linesplt.index("CDR1.IMGT.length")
53 			cdr2LengthIndex = linesplt.index("CDR2.IMGT.length")
54 			first = False
55 			continue
56 		linecount += 1
57 		linesplt = line.split("\t")
58 		ID = linesplt[IDIndex]
59 		genedic[ID] = linesplt[best_matchIndex]
60 		try:
61 			mutationdic[ID + "_FR1"] = [mutationMatcher.match(x).groups() for x in linesplt[fr1Index].split("|") if x] if (linesplt[fr1Index] != "NA" and empty_region_filter == "leader") else []
62 			mutationdic[ID + "_CDR1"] = [mutationMatcher.match(x).groups() for x in linesplt[cdr1Index].split("|") if x] if (linesplt[cdr1Index] != "NA" and empty_region_filter in ["leader", "FR1"]) else []
63 			mutationdic[ID + "_FR2"] = [mutationMatcher.match(x).groups() for x in linesplt[fr2Index].split("|") if x] if (linesplt[fr2Index] != "NA" and empty_region_filter in ["leader", "FR1", "CDR1"]) else []
64 			mutationdic[ID + "_CDR2"] = [mutationMatcher.match(x).groups() for x in linesplt[cdr2Index].split("|") if x] if (linesplt[cdr2Index] != "NA") else []
65 			mutationdic[ID + "_FR2-CDR2"] = mutationdic[ID + "_FR2"] + mutationdic[ID + "_CDR2"]
66 			mutationdic[ID + "_FR3"] = [mutationMatcher.match(x).groups() for x in linesplt[fr3Index].split("|") if x] if linesplt[fr3Index] != "NA" else []
67 		except Exception as e:
68 			print "Something went wrong while processing this line:"
69 			print linesplt
70 			print linecount
71 			print e
72 		mutationList += mutationdic[ID + "_FR1"] + mutationdic[ID + "_CDR1"] + mutationdic[ID + "_FR2"] + mutationdic[ID + "_CDR2"] + mutationdic[ID + "_FR3"]
73 		mutationListByID[ID] = mutationdic[ID + "_FR1"] + mutationdic[ID + "_CDR1"] + mutationdic[ID + "_FR2"] + mutationdic[ID + "_CDR2"] + mutationdic[ID + "_FR3"]
74 
75 		cdr1Length = linesplt[cdr1LengthIndex]
76 		cdr2Length = linesplt[cdr2LengthIndex]
77 
78 		cdr1LengthDic[ID] = int(cdr1Length) if cdr1Length != "X" else 0
79 		cdr2LengthDic[ID] = int(cdr2Length) if cdr2Length != "X" else 0
80 		
81 		IDlist += [ID]
82 
83 AALength = (int(max(mutationList, key=lambda i: int(i[4]) if i[4] else 0)[4]) + 1)  # [4] is the position of the AA mutation, None if silent
84 if AALength < 60:
85 	AALength = 64
86 
87 AA_mutation = [0] * AALength
88 AA_mutation_dic = {"IGA": AA_mutation[:], "IGG": AA_mutation[:], "IGM": AA_mutation[:], "IGE": AA_mutation[:], "unm": AA_mutation[:], "all": AA_mutation[:]}
89 AA_mutation_empty = AA_mutation[:]
90 
91 aa_mutations_by_id_file = outfile[:outfile.rindex("/")] + "/aa_id_mutations.txt"
92 with open(aa_mutations_by_id_file, 'w') as o:
93 	o.write("ID\tbest_match\t" + "\t".join([str(x) for x in range(1,AALength)]) + "\n")
94 	for ID in mutationListByID.keys():
95 		AA_mutation_for_ID = AA_mutation_empty[:]
96 		for mutation in mutationListByID[ID]:
97 			if mutation[4]:
98 				AA_mutation_position = int(mutation[4])
99 				AA_mutation[AA_mutation_position] += 1
00 				AA_mutation_for_ID[AA_mutation_position] += 1
01 				clss = genedic[ID][:3]
02 				AA_mutation_dic[clss][AA_mutation_position] += 1
03 		o.write(ID + "\t" + genedic[ID] + "\t" + "\t".join([str(x) for x in AA_mutation_for_ID[1:]]) + "\n")
04 
05 
06 
07 #absent AA stuff
08 absentAACDR1Dic = defaultdict(list)
09 absentAACDR1Dic[5] = range(29,36)
10 absentAACDR1Dic[6] = range(29,35)
11 absentAACDR1Dic[7] = range(30,35)
12 absentAACDR1Dic[8] = range(30,34)
13 absentAACDR1Dic[9] = range(31,34)
14 absentAACDR1Dic[10] = range(31,33)
15 absentAACDR1Dic[11] = [32]
16 
17 absentAACDR2Dic = defaultdict(list)
18 absentAACDR2Dic[0] = range(55,65)
19 absentAACDR2Dic[1] = range(56,65)
20 absentAACDR2Dic[2] = range(56,64)
21 absentAACDR2Dic[3] = range(57,64)
22 absentAACDR2Dic[4] = range(57,63)
23 absentAACDR2Dic[5] = range(58,63)
24 absentAACDR2Dic[6] = range(58,62)
25 absentAACDR2Dic[7] = range(59,62)
26 absentAACDR2Dic[8] = range(59,61)
27 absentAACDR2Dic[9] = [60]
28 
29 absentAA = [len(IDlist)] * (AALength-1)
30 for k, cdr1Length in cdr1LengthDic.iteritems():
31 	for c in absentAACDR1Dic[cdr1Length]:
32 		absentAA[c] -= 1
33 
34 for k, cdr2Length in cdr2LengthDic.iteritems():
35 	for c in absentAACDR2Dic[cdr2Length]:
36 		absentAA[c] -= 1
37 
38 
39 aa_mutations_by_id_file = outfile[:outfile.rindex("/")] + "/absent_aa_id.txt"
40 with open(aa_mutations_by_id_file, 'w') as o:
41 	o.write("ID\tcdr1length\tcdr2length\tbest_match\t" + "\t".join([str(x) for x in range(1,AALength)]) + "\n")
42 	for ID in IDlist:
43 		absentAAbyID = [1] * (AALength-1)
44 		cdr1Length = cdr1LengthDic[ID]
45 		for c in absentAACDR1Dic[cdr1Length]:
46 			absentAAbyID[c] -= 1
47 
48 		cdr2Length = cdr2LengthDic[ID]
49 		for c in absentAACDR2Dic[cdr2Length]:
50 			absentAAbyID[c] -= 1
51 		o.write(ID + "\t" + str(cdr1Length) + "\t" + str(cdr2Length) + "\t" + genedic[ID] + "\t" + "\t".join([str(x) for x in absentAAbyID]) + "\n")
52 
53 if linecount == 0:
54 	print "No data, exiting"
55 	with open(outfile, 'w') as o:
56 		o.write("RGYW (%)," + ("0,0,0\n" * len(genes)))
57 		o.write("WRCY (%)," + ("0,0,0\n" * len(genes)))
58 		o.write("WA (%)," + ("0,0,0\n" * len(genes)))
59 		o.write("TW (%)," + ("0,0,0\n" * len(genes)))
60 	import sys
61 
62 	sys.exit()
63 
64 hotspotMatcher = re.compile("[actg]+,(\d+)-(\d+)\((.*)\)")
65 RGYWCount = {}
66 WRCYCount = {}
67 WACount = {}
68 TWCount = {}
69 
70 #IDIndex = 0
71 ataIndex = 0
72 tatIndex = 0
73 aggctatIndex = 0
74 atagcctIndex = 0
75 first = True
76 with open(infile, 'r') as i:
77 	for line in i:
78 		if first:
79 			linesplt = line.split("\t")
80 			ataIndex = linesplt.index("X.a.t.a")
81 			tatIndex = linesplt.index("t.a.t.")
82 			aggctatIndex = linesplt.index("X.a.g.g.c.t..a.t.")
83 			atagcctIndex = linesplt.index("X.a.t..a.g.c.c.t.")
84 			first = False
85 			continue
86 		linesplt = line.split("\t")
87 		gene = linesplt[best_matchIndex]
88 		ID = linesplt[IDIndex]
89 		RGYW = [(int(x), int(y), z) for (x, y, z) in
90 				[hotspotMatcher.match(x).groups() for x in linesplt[aggctatIndex].split("|") if x]]
91 		WRCY = [(int(x), int(y), z) for (x, y, z) in
92 				[hotspotMatcher.match(x).groups() for x in linesplt[atagcctIndex].split("|") if x]]
93 		WA = [(int(x), int(y), z) for (x, y, z) in
94 			  [hotspotMatcher.match(x).groups() for x in linesplt[ataIndex].split("|") if x]]
95 		TW = [(int(x), int(y), z) for (x, y, z) in
96 			  [hotspotMatcher.match(x).groups() for x in linesplt[tatIndex].split("|") if x]]
97 		RGYWCount[ID], WRCYCount[ID], WACount[ID], TWCount[ID] = 0, 0, 0, 0
98 
99 		mutationList = mutationdic[ID + "_FR1"] + mutationdic[ID + "_CDR1"] + mutationdic[ID + "_FR2"] + mutationdic[ID + "_CDR2"] + mutationdic[ID + "_FR3"]
00 		for mutation in mutationList:
01 			frm, where, to, AAfrm, AAwhere, AAto, junk = mutation
02 			mutation_in_RGYW = any([(start <= int(where) <= end) for (start, end, region) in RGYW])
03 			mutation_in_WRCY = any([(start <= int(where) <= end) for (start, end, region) in WRCY])
04 			mutation_in_WA = any([(start <= int(where) <= end) for (start, end, region) in WA])
05 			mutation_in_TW = any([(start <= int(where) <= end) for (start, end, region) in TW])
06 
07 			in_how_many_motifs = sum([mutation_in_RGYW, mutation_in_WRCY, mutation_in_WA, mutation_in_TW])
08 
09 			if in_how_many_motifs > 0:
10 				RGYWCount[ID] += (1.0 * int(mutation_in_RGYW)) / in_how_many_motifs
11 				WRCYCount[ID] += (1.0 * int(mutation_in_WRCY)) / in_how_many_motifs
12 				WACount[ID] += (1.0 * int(mutation_in_WA)) / in_how_many_motifs
13 				TWCount[ID] += (1.0 * int(mutation_in_TW)) / in_how_many_motifs
14 
15 
16 def mean(lst):
17 	return (float(sum(lst)) / len(lst)) if len(lst) > 0 else 0.0
18 
19 
20 def median(lst):
21 	lst = sorted(lst)
22 	l = len(lst)
23 	if l == 0:
24 		return 0
25 	if l == 1:
26 		return lst[0]
27 		
28 	l = int(l / 2)
29 	
30 	if len(lst) % 2 == 0:
31 		return float(lst[l] + lst[(l - 1)]) / 2.0
32 	else:
33 		return lst[l]
34 
35 funcs = {"mean": mean, "median": median, "sum": sum}
36 
37 directory = outfile[:outfile.rfind("/") + 1]
38 value = 0
39 valuedic = dict()
40 
41 for fname in funcs.keys():
42 	for gene in genes:
43 		with open(directory + gene + "_" + fname + "_value.txt", 'r') as v:
44 			valuedic[gene + "_" + fname] = float(v.readlines()[0].rstrip())
45 	with open(directory + "all_" + fname + "_value.txt", 'r') as v:
46 		valuedic["total_" + fname] = float(v.readlines()[0].rstrip())
47 	
48 
49 def get_xyz(lst, gene, f, fname):
50 	x = round(round(f(lst), 1))
51 	y = valuedic[gene + "_" + fname]
52 	z = str(round(x / float(y) * 100, 1)) if y != 0 else "0"
53 	return (str(x), str(y), z)
54 
55 dic = {"RGYW": RGYWCount, "WRCY": WRCYCount, "WA": WACount, "TW": TWCount}
56 arr = ["RGYW", "WRCY", "WA", "TW"]
57 
58 geneMatchers = {gene: re.compile("^" + gene + ".*") for gene in genes}
59 
60 for fname in funcs.keys():
61 	func = funcs[fname]
62 	foutfile = outfile[:outfile.rindex("/")] + "/hotspot_analysis_" + fname + ".txt"
63 	with open(foutfile, 'w') as o:
64 		for typ in arr:
65 			o.write(typ + " (%)")
66 			curr = dic[typ]
67 			for gene in genes:
68 				geneMatcher = geneMatchers[gene]
69 				if valuedic[gene + "_" + fname] is 0:
70 					o.write(",0,0,0")
71 				else:
72 					x, y, z = get_xyz([curr[x] for x in [y for y, z in genedic.iteritems() if geneMatcher.match(z)]], gene, func, fname)
73 					o.write("," + x + "," + y + "," + z)
74 			x, y, z = get_xyz([y for x, y in curr.iteritems() if not genedic[x].startswith("unmatched")], "total", func, fname)
75 			#x, y, z = get_xyz([y for x, y in curr.iteritems()], "total", func, fname)
76 			o.write("," + x + "," + y + "," + z + "\n")
77 
78 
79 # for testing
80 seq_motif_file = outfile[:outfile.rindex("/")] + "/motif_per_seq.txt"
81 with open(seq_motif_file, 'w') as o:
82 	o.write("ID\tRGYW\tWRCY\tWA\tTW\n")
83 	for ID in IDlist:
84 		#o.write(ID + "\t" + str(round(RGYWCount[ID], 2)) + "\t" + str(round(WRCYCount[ID], 2)) + "\t" + str(round(WACount[ID], 2)) + "\t" + str(round(TWCount[ID], 2)) + "\n")
85 		o.write(ID + "\t" + str(RGYWCount[ID]) + "\t" + str(WRCYCount[ID]) + "\t" + str(WACount[ID]) + "\t" + str(TWCount[ID]) + "\n")