Mercurial > repos > genouest > get_pairs

comparison get_pairs.py @ 0:ac738de70427 draft default tip

Find changesets by keywords (author, files, the commit message), revision number or hash, or revset expression.
planemo upload for repository https://github.com/genouest/galaxy-tools/tree/master/tools/get_pairs commit 02689a4f03aff06ce7f56ffd329ad72befdc692c
author genouest
date Mon, 10 Sep 2018 10:16:42 -0400
parents
children
comparison
equal deleted inserted replaced
-1:000000000000 0:ac738de70427
1 #!/opt/python/bin/python
2 # -*- coding: utf-8 -*-
3 # ----------------------------------------------------------
4 # --
5 # -- author : Pierre Pericard
6 # -- created : 2012-11-09
7 # -- modified: 2013-05-23
8 # --
9 # ----------------------------------------------------------
10 # --
11 # -- description : Get separately paired reads and singletons
12 # -- from two fastq files (left and right)
13 # --
14 # -- get_pairs.py file1.fastq file2.fastq
15 # --
16 # ----------------------------------------------------------
17
18 import argparse
19 import sys
20
21
22 if __name__ == '__main__':
23
24 # Arguments
25 parser = argparse.ArgumentParser(description='Get separately paired reads and singletons from two fastq files (left and right)')
26 parser.add_argument('leftreads', metavar='leftreads', type=argparse.FileType('r'), help='left reads fastq')
27 parser.add_argument('rightreads', metavar='rightreads', type=argparse.FileType('r'), help='right reads fastq')
28
29 args = parser.parse_args()
30
31 leftreads = args.leftreads.name
32 rightreads = args.rightreads.name
33
34 (n1, n2) = (list(), list())
35
36 for f, n in ((leftreads, n1), (rightreads, n2)):
37 with open(f, 'r') as fh:
38 c = 0
39 for line in fh:
40 line = line.strip()
41 if line:
42 c += 1
43 if c % 4 == 1:
44 n.append(line.split()[0][1:].split('/')[0])
45 if c % 40000 == 1:
46 sys.stdout.write("\r%.2f M reads read" % (c / 4000000.0))
47 sys.stdout.write("\r%.2f M reads read\n" % (c / 4000000.0))
48
49 notcommon = set(n1) ^ set(n2)
50
51 for f in (leftreads, rightreads):
52
53 if f == leftreads:
54 basefilename = "left"
55 else:
56 basefilename = "right"
57
58 pfh = open(basefilename + '.paired.fastq', 'w')
59 ufh = open(basefilename + '.unpaired.fastq', 'w')
60 with open(f, 'r') as fh:
61 c = 0
62 paired = False
63 for line in fh:
64 line = line.strip()
65 if line:
66 c += 1
67 if c % 4 == 1:
68 paired = line.split()[0][1:].split('/')[0] not in notcommon
69 if c % 40000 == 1:
70 sys.stdout.write("\r%.2f M reads writen" % (c / 4000000.0))
71 if paired:
72 pfh.write("%s\n" % line)
73 else:
74 ufh.write("%s\n" % line)
75 sys.stdout.write("\r%.2f M reads writen\n" % (c / 4000000.0))
76 pfh.close()
77 ufh.close()