-
Notifications
You must be signed in to change notification settings - Fork 0
/
eval.py
461 lines (446 loc) · 20.5 KB
/
eval.py
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
346
347
348
349
350
351
352
353
354
355
356
357
358
359
360
361
362
363
364
365
366
367
368
369
370
371
372
373
374
375
376
377
378
379
380
381
382
383
384
385
386
387
388
389
390
391
392
393
394
395
396
397
398
399
400
401
402
403
404
405
406
407
408
409
410
411
412
413
414
415
416
417
418
419
420
421
422
423
424
425
426
427
428
429
430
431
432
433
434
435
436
437
438
439
440
441
442
443
444
445
446
447
448
449
450
451
452
453
454
455
456
457
458
459
460
461
from pysam import VariantFile
import vcf
import sys
import os
def parse_to_int(sth):
if sth == None:
return 0
elif isinstance(sth, str):
return int(sth)
elif isinstance(sth, list):
return parse_to_int(sth[0])
elif isinstance(sth, tuple):
return parse_to_int(sth[0])
elif isinstance(sth, int):
return sth
else:
return sth
def phase_bnd(alt, chr, pos):
# print(alt)
# print(str(alt[0]))
if alt[0] == ']':
form = ']]N'
chr2 = alt.split(':')[0][1:]
pos2 = int(alt.split(':')[1][:-2])
elif alt[0] == '[':
form = '[[N'
chr2 = alt.split(':')[0][1:]
pos2 = int(alt.split(':')[1][:-2])
else:
# print(type(alt))
if alt[1] == ']':
form = 'N]]'
chr2 = alt.split(':')[0][2:]
pos2 = int(alt.split(':')[1][:-1])
else:
form = 'N[['
chr2 = alt.split(':')[0][2:]
pos2 = int(alt.split(':')[1][:-1])
try:
if int(chr) <= int(chr2):
if form == 'N[[':
form = ']]N'
if form == ']]N':
form = 'N[['
return form, chr, pos, chr2, pos2
else:
return form, chr2, pos2, chr, pos
except:
return form, chr, pos, chr2, pos2
# end：end[TRA], svlen[OTHERS]
def parse_record(record):
sample_ids = ['CHM13', 'CHM1', 'HG00268', 'HG00514', 'HG00733', 'HG01352', 'HG02059', 'HG02106', 'HG02818', 'HG04217', 'HX1', 'NA12878', 'NA19240', 'NA19434']
sv_type = record.info['SVTYPE'][:3]
if sv_type == 'TRA':
sv_type = 'BND'
chrom1 = record.chrom
chrom2 = record.chrom
start = parse_to_int(record.pos)
try:
svlen = record.info['SVLEN']
try:
if len(svlen):
svlen = svlen[0]
except:
pass
except:
svlen = 0
try:
end = record.stop
except:
end = start
if sv_type == 'BND':
bnd = str(record.alts[0])
form, chrom1, start, chrom2, end = phase_bnd(bnd, chrom1, start)
supp_id = record.info['SUPP_VEC']
supp_vec = list()
for i in range(len(supp_id)):
if supp_id[i] == '1':
supp_vec.append(sample_ids[i])
return sv_type, chrom1, start, end, abs(svlen), supp_vec, chrom2
def check_same_variant(sv_type, start1, start2, end1, end2, len1, len2):
#print('%s, %d, %d, %d, %d'%(sv_type, start1, start2, end1, end2))
if sv_type == 'INS':
if abs(start1 - start2) < 1000 and 0.7 < min(len1, len2) / max(len1, len2) <= 1:
return True
else:
if abs(start1 - start2) < 1000 and 0.7 < min(len1, len2) / max(len1, len2) <= 1: # and max(start1, start2) <= min(end1, end2):
return True
return False
def check_same_bnd(sv_type, start1, start2, end1, end2):
if abs(start1 - start2) < 1000 and abs(end1 - end2) < 1000:
return True
return False
def parse_vcf(filename):
vcf_reader = VariantFile(filename, 'r')
sv_dict = dict()
for record in vcf_reader.fetch():
sv_type, chrom1, start, end, svlen, supp_id, chrom2 = parse_record(record)
if sv_type not in sv_dict:
sv_dict[sv_type] = dict()
if sv_type == 'BND':
if chrom1 not in sv_dict[sv_type]:
sv_dict[sv_type][chrom1] = dict()
if chrom2 not in sv_dict[sv_type][chrom1]:
sv_dict[sv_type][chrom1][chrom2] = list()
sv_dict[sv_type][chrom1][chrom2].append([start, end, svlen, supp_id])
else:
if chrom1 in sv_dict[sv_type]:
sv_dict[sv_type][chrom1].append([start, end, svlen, supp_id])
else:
sv_dict[sv_type][chrom1] = [[start, end, svlen, supp_id]]
return sv_dict
def parse_chr_merge(filename):
sample_ids = ['CHM1', 'CHM13', 'HG00268', 'HG00514', 'HG00733', 'HG01352', 'HG02059', 'HG02106', 'HG02818', 'HG04217', 'HX1', 'NA12878', 'NA19240', 'NA19434']
sv_dict = dict()
with open(filename, 'r') as f:
for line in f:
if line[:3] == 'CHM':
continue
lineseq = line.strip().split('\t')
chrom1 = lineseq[0]
start = int(lineseq[1])
sv_type = lineseq[2].split('SVTYPE=')[1].split(';')[0]
if sv_type not in sv_dict:
sv_dict[sv_type] = dict()
svlen = int(lineseq[2].split(';SVLEN=')[1].split(';')[0])
end = int(lineseq[2].split(';END=')[1].split(';')[0])
if sv_type == 'INV':
svlen = end - start + 1
samples = list()
for i in range(14):
if lineseq[i + 3] != '0/0':
samples.append(sample_ids[i])
if chrom1 in sv_dict[sv_type]:
sv_dict[sv_type][chrom1].append([start, end, svlen, samples])
else:
sv_dict[sv_type][chrom1] = [[start, end, svlen, samples]]
return sv_dict
# sample_flag == 1: take sample into account
def solve_two_vcf(sv_dict1, sv_dict2, output, sample_flag): # sv_dict1是被比较的
file = open(output, 'a')
right = dict()
wrong = dict()
vis = set()
idx = 0
for sv_type in ['INS', 'DEL', 'INV', 'DUP', 'BND']:
right[sv_type] = 0
wrong[sv_type] = 0
for sv_type in ['INS', 'DEL', 'INV', 'DUP']:
#for sv_type in ['INV']:
if sv_type not in sv_dict1:
continue
for chrom in sv_dict1[sv_type].keys():
for record_list in sv_dict1[sv_type][chrom]:
'''
if sv_type == 'INV':
print('==')
print(record_list)
idx += 1
'''
flag = 0
if chrom in sv_dict2[sv_type]:
for temp_list in sv_dict2[sv_type][chrom]:
'''
if record_list[0] == 64704785 and temp_list[0] == 64704786:
print(record_list[1])
print(temp_list[1])
print(check_same_variant(sv_type, record_list[0], temp_list[0], record_list[1], temp_list[1]))
'''
if check_same_variant(sv_type, record_list[0], temp_list[0], record_list[1], temp_list[1], record_list[2], temp_list[2]):
if sample_flag == 0:
file.write('same variant on chr%s (%d, %d, %d), (%d, %d, %d)\n'%(chrom, record_list[0], record_list[1], record_list[2], temp_list[0], temp_list[1], temp_list[2]))
flag = 1
right[sv_type] += 1
else:
for sample_id in record_list[2]:
if sample_id in temp_list[2]:
right[sv_type] += 1
break
if flag == 0:
file.write('wrong on chr%s %s, (%d, %d, %d)\n'%(chrom, sv_type, record_list[0], record_list[1], record_list[2]))
if sample_flag == 0:
wrong[sv_type] += 1
else:
wrong[sv_type] += len(record_list[2])
#if idx == 8:
# exit(0)
if 'BND' in sv_dict1:
sv_type = 'BND'
for chrom in sv_dict1[sv_type].keys():
for chrom2 in sv_dict1[sv_type][chrom].keys():
for record_list in sv_dict1[sv_type][chrom][chrom2]:
flag = 0
if sv_type in sv_dict2 and chrom in sv_dict2[sv_type] and chrom2 in sv_dict2[sv_type][chrom]:
for temp_list in sv_dict2[sv_type][chrom][chrom2]:
if check_same_bnd(sv_type, record_list[0], temp_list[0], record_list[1], temp_list[1]):
file.write('same variant on chr(%s, %d, %d), (%s, %d, %d)\n'%(chrom, record_list[0], record_list[1], chrom2, temp_list[0], temp_list[1]))
flag = 1
right[sv_type] += 1
break
if flag == 0:
file.write('wrong on chr%s %s %s, (%d, %d)\n'%(chrom, chrom2, sv_type, record_list[0], record_list[1]))
wrong[sv_type] += 1
return right, wrong
def solve_tp(sv_dict1, sv_dict2): # sv_dict1是被比较的
right = dict()
wrong = dict()
vis = set()
idx = 0
tp = list()
fx = list()
for sv_type in ['INS', 'DEL', 'INV', 'DUP', 'BND']:
right[sv_type] = 0
wrong[sv_type] = 0
for sv_type in ['INS', 'DEL', 'INV', 'DUP']:
#for sv_type in ['INV']:
if sv_type not in sv_dict1:
continue
for chrom in sv_dict1[sv_type].keys():
for record_list in sv_dict1[sv_type][chrom]:
flag = 0
if chrom in sv_dict2[sv_type]:
for temp_list in sv_dict2[sv_type][chrom]:
if check_same_variant(sv_type, record_list[0], temp_list[0], record_list[1], temp_list[1], record_list[2], temp_list[2]):
#file.write('same variant on chr%s (%d, %d, %d), (%d, %d, %d)\n'%(chrom, record_list[0], record_list[1], record_list[2], temp_list[0], temp_list[1], temp_list[2]))
tp.append([chrom, record_list[0], sv_type, record_list[1], record_list[2]]) # chrom, pos, type, end, len
flag = 1
right[sv_type] += 1
break
if flag == 0:
#file.write('wrong on chr%s %s, (%d, %d, %d)\n'%(chrom, sv_type, record_list[0], record_list[1], record_list[2]))
fx.append([chrom, record_list[0], sv_type, record_list[1], record_list[2]])
wrong[sv_type] += 1
#if idx == 8:
# exit(0)
if 'BND' in sv_dict1:
sv_type = 'BND'
for chrom in sv_dict1[sv_type].keys():
for chrom2 in sv_dict1[sv_type][chrom].keys():
for record_list in sv_dict1[sv_type][chrom][chrom2]:
flag = 0
if sv_type in sv_dict2 and chrom in sv_dict2[sv_type] and chrom2 in sv_dict2[sv_type][chrom]:
for temp_list in sv_dict2[sv_type][chrom][chrom2]:
if check_same_bnd(sv_type, record_list[0], temp_list[0], record_list[1], temp_list[1]):
#file.write('same variant on chr(%s, %d, %d), (%s, %d, %d)\n'%(chrom, record_list[0], record_list[1], chrom2, temp_list[0], temp_list[1]))
tp.append([chrom, record_list[0], sv_type, record_list[1], record_list[2]])
flag = 1
right[sv_type] += 1
break
if flag == 0:
#file.write('wrong on chr%s %s %s, (%d, %d)\n'%(chrom, chrom2, sv_type, record_list[0], record_list[1]))
fx.append([chrom, record_list[0], sv_type, record_list[1], record_list[2]])
wrong[sv_type] += 1
return right, wrong, tp, fx
if __name__ == '__main__':
if len(sys.argv) == 4:
vcf_standard = sys.argv[2]
vcf_test = sys.argv[3]
#sv_dict1 = parse_chr_merge(vcf_standard)
sv_dict1 = parse_vcf(vcf_standard)
sv_dict2 = parse_vcf(vcf_test)
output = sys.argv[1]
if not os.path.exists(output):
os.mkdir(output)
right1, wrong1, tpbase, fn = solve_tp(sv_dict1, sv_dict2) # standard_cnt
right2, wrong2, tpcall, fp = solve_tp(sv_dict2, sv_dict1) # test_cnt
#exit(0)
right1_cnt = 0
right2_cnt = 0
standard_cnt = 0
test_cnt = 0
#print(right2)
#print(wrong2)
print('SVTYPE\tprecision\trecall\t\tF1\t\tTP-base\tTP-call\tbase\tcall\t')
for sv_type in ['INS', 'DEL', 'INV', 'DUP', 'BND']:
#for sv_type in ['DEL']:
right1_cnt += right1[sv_type]
right2_cnt += right2[sv_type]
standard_cnt += right1[sv_type] + wrong1[sv_type]
test_cnt += right2[sv_type] + wrong2[sv_type]
precision = -1 if right2[sv_type] + wrong2[sv_type] == 0 else right2[sv_type] / (right2[sv_type] + wrong2[sv_type])
recall = -1 if right1[sv_type] + wrong1[sv_type] == 0 else right1[sv_type] / (right1[sv_type] + wrong1[sv_type])
F1 = 2 * precision * recall / (precision + recall)
print('%s\t%f\t%f\t%f\t%d\t%d\t%d\t%d'%(sv_type, round(precision, 6), round(recall, 6), round(F1, 6), right1[sv_type], right2[sv_type], right1[sv_type] + wrong1[sv_type], right2[sv_type] + wrong2[sv_type]))
precision = right2_cnt / test_cnt
recall = right1_cnt / standard_cnt
F1 = 2 * precision * recall / (precision + recall)
print('%s\t%f\t%f\t%f\t%d\t%d\t%d\t%d'%('TOTAL', round(precision, 6), round(recall, 6), round(F1, 6), right1_cnt, right2_cnt, standard_cnt, test_cnt))
# chrom, pos, type, end, len
tpbase = sorted(tpbase, key = lambda x:(x[0],x[1]))
tpcall = sorted(tpcall, key = lambda x:(x[0],x[1]))
fp = sorted(fp, key = lambda x:(x[0],x[1]))
fn = sorted(fn, key = lambda x:(x[0],x[1]))
with open(output + '/tp-base.txt', 'w') as f:
for x in tpbase:
f.write('%s\t%d\tSVTYPE=%s;END=%d;SVLEN=%d\n'%(x[0], x[1], x[2], x[3], x[4]))
with open(output + '/tp-call.txt', 'w') as f:
for x in tpcall:
f.write('%s\t%d\tSVTYPE=%s;END=%d;SVLEN=%d\n'%(x[0], x[1], x[2], x[3], x[4]))
with open(output + '/fp.txt', 'w') as f:
for x in fp:
f.write('%s\t%d\tSVTYPE=%s;END=%d;SVLEN=%d\n'%(x[0], x[1], x[2], x[3], x[4]))
with open(output + '/fn.txt', 'w') as f:
for x in fn:
f.write('%s\t%d\tSVTYPE=%s;END=%d;SVLEN=%d\n'%(x[0], x[1], x[2], x[3], x[4]))
print('finish eval')
elif len(sys.argv) == 5:
vcf_standard = sys.argv[1]
vcf_test = sys.argv[2]
sv_dict1 = parse_chr_merge(vcf_standard)
sv_dict2 = parse_vcf(vcf_test)
output = sys.argv[3]
file = open(output, 'w')
file.close()
sample_flag = int(sys.argv[4])
right1, wrong1 = solve_two_vcf(sv_dict1, sv_dict2, output, sample_flag) # standard_cnt
right2, wrong2 = solve_two_vcf(sv_dict2, sv_dict1, output, sample_flag) # test_cnt
#exit(0)
right1_cnt = 0
right2_cnt = 0
standard_cnt = 0
test_cnt = 0
#print(right2)
#print(wrong2)
print('SVTYPE\tprecision\trecall\t\tF1\t\tTP-base\tTP-call\tbase\tcall\t')
for sv_type in ['INS', 'DEL', 'INV', 'DUP', 'BND']:
#for sv_type in ['DEL']:
right1_cnt += right1[sv_type]
right2_cnt += right2[sv_type]
standard_cnt += right1[sv_type] + wrong1[sv_type]
test_cnt += right2[sv_type] + wrong2[sv_type]
precision = -1 if right2[sv_type] + wrong2[sv_type] == 0 else right2[sv_type] / (right2[sv_type] + wrong2[sv_type])
recall = -1 if right1[sv_type] + wrong1[sv_type] == 0 else right1[sv_type] / (right1[sv_type] + wrong1[sv_type])
F1 = 2 * precision * recall / (precision + recall)
print('%s\t%f\t%f\t%f\t%d\t%d\t%d\t%d'%(sv_type, round(precision, 6), round(recall, 6), round(F1, 6), right1[sv_type], right2[sv_type], right1[sv_type] + wrong1[sv_type], right2[sv_type] + wrong2[sv_type]))
precision = right2_cnt / test_cnt
recall = right1_cnt / standard_cnt
F1 = 2 * precision * recall / (precision + recall)
print('%s\t%f\t%f\t%f\t%d\t%d\t%d\t%d'%('TOTAL', round(precision, 6), round(recall, 6), round(F1, 6), right1_cnt, right2_cnt, standard_cnt, test_cnt))
elif len(sys.argv) == 1:
output = 'eval.tmp'
file = open(output, 'w')
file.close()
sv_dict1 = parse_chr_merge('chr1_merge.vcf')
#tool_list = ['output.bak311/merged', 'output322/merged', 'output.bak311/survivor']
tool_list = ['output0/merged']
#name_list = ['src', 'src1', 'sur']
name_list = ['src', 'sur']
for call_tool in ['cutesv', 'pbsv', 'sniffles', 'svim', 'cutesv0', 'svimS4']:
print('===========================' + call_tool + '===========================')
#print('SVTYPE\tprecision\trecall\t\tF1\t\tTP-base\tTP-call\tbase\tcall\t')
for i in range(1):
merge_tool = tool_list[i]
vcf_test = merge_tool + '_' + call_tool + '.vcf'
sv_dict2 = parse_vcf(vcf_test)
right1, wrong1 = solve_two_vcf(sv_dict1, sv_dict2, output, 0)
right2, wrong2 = solve_two_vcf(sv_dict2, sv_dict1, output, 0)
#print(right1)
#print(right2)
right1_cnt = 0
right2_cnt = 0
standard_cnt = 0
test_cnt = 0
print('SVTYPE\tprecision\trecall\t\tF1\t\tTP-base\tTP-call\tbase\tcall\t')
for sv_type in ['INS', 'DEL', 'INV', 'DUP']:
#for sv_type in ['DEL']:
right1_cnt += right1[sv_type]
right2_cnt += right2[sv_type]
standard_cnt += right1[sv_type] + wrong1[sv_type]
test_cnt += right2[sv_type] + wrong2[sv_type]
precision = -1 if right2[sv_type] + wrong2[sv_type] == 0 else right2[sv_type] / (right2[sv_type] + wrong2[sv_type])
recall = -1 if right1[sv_type] + wrong1[sv_type] == 0 else right1[sv_type] / (right1[sv_type] + wrong1[sv_type])
F1 = 2 * precision * recall / (precision + recall)
print('%s\t%f\t%f\t%f\t%d\t%d\t%d\t%d'%(sv_type, round(precision, 6), round(recall, 6), round(F1, 6), right1[sv_type], right2[sv_type], right1[sv_type] + wrong1[sv_type], right2[sv_type] + wrong2[sv_type]))
precision = right2_cnt / test_cnt
recall = right1_cnt / standard_cnt
F1 = 2 * precision * recall / (precision + recall)
print('%s\t%f\t%f\t%f\t%d\t%d\t%d\t%d'%('TOTAL', round(precision, 6), round(recall, 6), round(F1, 6), right1_cnt, right2_cnt, standard_cnt, test_cnt))
else:
print('parameter error %d'%(len(sys.argv)))
#'''
'''
def parse_vcf(filename):
vcf_reader = VariantFile(filename, 'r')
sv_dict = dict()
for bx in ['INS', 'DEL', 'INV', 'DUP', 'TRA']:
sv_dict[bx] = dict()
for record in vcf_reader.fetch():
sv_type, chrom1, chrom2, start, end, name = parse_record(record, 'NULL')
if sv_type == 'TRA':
if chrom1 in sv_dict[sv_type]:
if chrom2 in sv_dict[sv_type][chrom1]:
sv_dict[sv_type][chrom1][chrom2].append([start, end, name])
else:
sv_dict[sv_type][chrom1][chrom2] = [[start, end, name]]
else:
sv_dict[sv_type][chrom1] = dict()
sv_dict[sv_type][chrom1][chrom2] = [[start, end, name]]
else:
if chrom1 in sv_dict[sv_type]:
sv_dict[sv_type][chrom1].append([start, end, name])
else:
sv_dict[sv_type][chrom1] = [[start, end, name]]
return sv_dict
'''
def parse_nstd(filename):
sv_dict = dict()
for bx in ['INS', 'DEL', 'INV', 'DUP', 'BND']:
sv_dict[bx] = dict()
with open(filename, 'r') as f:
idx = 0
pres = 0
pree = 0
pre_type = ''
for line in f:
if line[0] == '#':
continue
lineseq = line.strip().split('\t')
chrom1 = lineseq[0]
if 'nstd162' in filename and chrom1 != '1':
break
start = int(lineseq[1])
sv_type = lineseq[7].split('SVTYPE=')[1].split(';')[0]
if sv_type == 'TRA':
sv_type = 'BND'
if 'DUP' in sv_type:
sv_type = 'DUP'
idx += 1
if sv_type == 'INS' or sv_type == 'DEL':
end = abs(int(lineseq[7].split(';SVLEN=')[1].split(';')[0]))
elif sv_type == 'BND':
end = 0
else:
end = abs(int(lineseq[7].split(';END=')[1].split(';')[0]))
name = 'variant' + str(idx)
#sample = lineseq[7].split('SAMPLE=')[1].split(';')[0]
if start == pres and pree == end and sv_type == pre_type and sv_type != 'BND':
continue
pres = start
pree = end
pre_type = sv_type
if chrom1 in sv_dict[sv_type]:
sv_dict[sv_type][chrom1].append([start, end, name])
else:
sv_dict[sv_type][chrom1] = [[start, end, name]]
return sv_dict