-
Notifications
You must be signed in to change notification settings - Fork 0
/
stat_target_region_from_bam_of_star.py
199 lines (178 loc) · 8.05 KB
/
stat_target_region_from_bam_of_star.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
import os
from glob import glob
import json
import plotly.graph_objs as go
from plotly.offline import plot as plt
import pandas as pd
import subprocess
def parse_star_final_out_log(logfiles:list, outdir=None):
results = list()
for logfile in logfiles:
if not outdir:
outdir = os.path.dirname(logfile)
sample = os.path.basename(logfile).split('.', 1)[0]
out_table = os.path.join(outdir, '{}.alignment_summary.json'.format(sample))
result = dict()
with open(logfile) as f:
for line in f:
if line.strip().endswith(":") or not line.strip():
continue
desc, data = line.split('|')
if "Number of input reads" in desc:
result['total_reads'] = int(data.strip())
elif 'Average input read length' in desc:
result['average_read_length'] = float(data.strip())
elif 'Uniquely mapped reads number' in desc:
result['unique_mapped'] = int(data.strip())
elif 'Uniquely mapped reads %' in desc:
result['unique_mapped_ratio'] = float(data.strip().strip('%'))
elif 'Average mapped length' in desc:
result['average_mapped_read_length'] = float(data.strip())
elif 'Number of splices: Total' in desc:
result['total_splices'] = int(data.strip())
elif 'Number of splices: Annotated' in desc:
result['annotated_splices'] = int(data.strip())
elif 'Number of reads mapped to multiple loci' in desc:
result['multiple_mapped'] = int(data.strip())
elif '% of reads mapped to multiple loci' in desc:
result['multiple_mapped_ratio'] = float(data.strip().strip('%'))
elif 'Number of reads mapped to too many loci' in desc:
result['many_mapped(>10)'] = int(data.strip())
elif '% of reads mapped to too many loci' in desc:
result['many_mapped_ratio'] = float(data.strip().strip('%'))
elif ' Number of chimeric reads' in desc:
result['chimeric'] = int(data.strip())
else:
pass
result['mapped_ratio'] = result['unique_mapped_ratio'] + result['multiple_mapped_ratio'] + result['many_mapped_ratio']
result['mapped'] = result['unique_mapped'] + result['multiple_mapped'] + result['many_mapped(>10)']
result['sample'] = sample
with open(out_table, 'w') as f:
json.dump(result, f, indent=2)
results.append(result)
pd.DataFrame(results).set_index('sample').to_csv('star.mapping.stat.csv')
return results
def parse_samtools_flagstat_result(stat_file):
result_list = list()
with open(stat_file) as f:
for line in f:
result_list.append(line.split(' ', 3))
return result_list
def parse_samtools_depth_result(stat_file, cov_limit=2000, step=10, outdir=None):
depth_dict = dict()
with open(stat_file) as f:
for line in f:
chr_, pos, depth = line.strip().split()
depth_dict.setdefault(int(depth), 0)
depth_dict[int(depth)] += 1
merge_dict = {'0': depth_dict[0]}
# print(sum(depth_dict.values()))
max_depth = max(depth_dict.keys())
if cov_limit >= max_depth:
cov_limit = max_depth
start = 0
while start < cov_limit:
start += step
key = '{}-{}'.format(start-step+1, start)
merge_dict.setdefault(key, 0)
for ind in range(start-step+1, start+1):
if ind in depth_dict:
merge_dict[key] += depth_dict.pop(ind)
key = '{}-{}'.format(start+1, max_depth)
merge_dict.setdefault(key, 0)
for ind in range(start+1, max_depth+1):
if ind in depth_dict:
merge_dict[key] += depth_dict.pop(ind)
if not outdir:
outdir = os.path.dirname(stat_file)
sample = os.path.basename(stat_file).split('.', 1)[0]
out_name = os.path.join(outdir, '{}.pos.depth.distribution.json'.format(sample))
with open(out_name, 'w') as f:
json.dump(merge_dict, f, indent=2)
return merge_dict
def stat_target_bam(bed, bam, rRNA_bed=None, overlap=0.05, rRNA_overlap=0.6, bedtools='bedtools', samtools='samtools',
outdir=None, threads=6, cov_limit=5000, step=10):
if not outdir:
outdir = os.path.dirname(bam)
if not outdir:
outdir = '.'
sample = os.path.basename(bam).split('.', 1)[0]
out = '{}/{}.target_region_bam.stat.txt'.format(outdir, sample)
cmd = '{} intersect '.format(bedtools)
cmd += '-a {} '.format(bam)
cmd += '-b {} '.format(bed)
cmd += '-wa '
cmd += '-F {} '.format(overlap)
cmd += '| '
cmd += '{} flagstat '.format(samtools)
cmd += '--threads {} '.format(threads/2)
cmd += '- '
cmd += '> {} '.format(out)
print(cmd)
p1 = subprocess.Popen(cmd, shell=True)
# get rRNA bam and stat
if rRNA_bed:
out_rRNA = '{}/{}.rRNA_bam.stat.txt'.format(outdir, sample)
cmd = '{} intersect '.format(bedtools)
cmd += '-a {} '.format(bam)
cmd += '-b {} '.format(rRNA_bed)
cmd += '-wa -s '
cmd += '-F {} '.format(rRNA_overlap)
cmd += '| '
cmd += '{} flagstat '.format(samtools)
cmd += '--threads {} '.format(threads/2)
cmd += '- '
cmd += '> {} '.format(out_rRNA)
print(cmd)
p2 = subprocess.Popen(cmd, shell=True)
# run samtools depth
depth_stat = os.path.join(outdir, '{}.pos.depth'.format(sample))
cmd = '{} depth '.format(samtools)
cmd += '-a '
cmd += '-d {} '.format(cov_limit)
cmd += '-b {} '.format(bed)
cmd += '{} '.format(bam)
cmd += '> {} '.format(depth_stat)
print(cmd)
subprocess.check_call(cmd, shell=True)
if p1.wait() != 0:
raise Exception("Failed to split out target bam based on bed!")
if rRNA_bed:
if p2.wait() != 0:
raise Exception("Failed to split out rRNA bam based on bed!")
chr_stat = os.path.join(outdir, '{}.chromosome.alignment.stat.txt'.format(sample))
cmd = '{} idxstats {} > {}'.format(samtools, bam, chr_stat)
subprocess.check_call(cmd, shell=True)
final_log = glob(os.path.join(os.path.dirname(bam), '*.Log.final.out'))[0]
original_bam_summary = parse_star_final_out_log([final_log], outdir=outdir)[0]
target_bam_summary = parse_samtools_flagstat_result(out)
target_ratio = int(target_bam_summary[4][0]) / original_bam_summary['mapped'] / 2
depth_dict = parse_samtools_depth_result(depth_stat, cov_limit=cov_limit, step=step)
if rRNA_bed:
rRNA_bam_summary = parse_samtools_flagstat_result(out_rRNA)
rRNA_ratio = int(rRNA_bam_summary[4][0]) / original_bam_summary['mapped'] / 2
original_bam_summary['rRNA_ratio'] = round(rRNA_ratio * 100, 2)
summary = os.path.join(outdir, '{}.target_region_alignment.summary.txt'.format(sample))
with open(summary, 'w') as f:
f.write('reads mapped to target region : {}/{}={:.2%}\n'.format(
target_bam_summary[4][0],
original_bam_summary['mapped']*2,
target_ratio)
)
target_region_length = sum(depth_dict.values())
none_zero_depth_region_length = target_region_length - depth_dict['0']
cover_ratio = none_zero_depth_region_length/target_region_length
f.write('non-zero depth target region: {}/{}={:.2%}\n'.format(
none_zero_depth_region_length,
target_region_length,
cover_ratio
))
original_bam_summary['target_capture_ratio'] = round(target_ratio*100, 2)
original_bam_summary['target_cover_ratio'] = round(cover_ratio*100, 2)
out_table = os.path.join(outdir, '{}.alignment_summary.json'.format(sample))
with open(out_table, 'w') as f:
json.dump(original_bam_summary, f, indent=2)
return target_ratio, cover_ratio
if __name__ == '__main__':
from xcmds import xcmds
xcmds.xcmds(locals(), include=['parse_star_final_out_log', 'stat_target_bam'])