Mercurial > repos > iuc > ivar_trim

comparison prepare_amplicon_info.py @ 5:cf65217ad61c draft

Find changesets by keywords (author, files, the commit message), revision number or hash, or revset expression.
"planemo upload for repository https://github.com/galaxyproject/tools-iuc/tree/master/tools/ivar/ commit a5ff06c631a2a5a0d5d44edd6cb58a599d50918b"
author iuc
date Wed, 19 May 2021 16:51:27 +0000
parents
children 5671e1d3d5ee
comparison
equal deleted inserted replaced
4:db536ad45f28 5:cf65217ad61c
1 #!/usr/bin/env python
2
3 # extends ivar trim's amplicon info parsing abilities
4 # to include calculation of amplicon regions from
5 # sets of nested (more than two) primers
6
7 import sys
8
9
10 # parse primers and their start positions from BED file
11 primer_starts = {}
12 with open(sys.argv[1]) as i:
13 for line in i:
14 f = line.strip().split('\t')
15 try:
16 if f[5] == '+':
17 primer_starts[f[3]] = int(f[1])
18 elif f[5] == '-':
19 primer_starts[f[3]] = int(f[2]) - 1
20 else:
21 raise ValueError()
22 except (IndexError, ValueError):
23 sys.exit(
24 'Primer BED file needs to be TAB-separated with the '
25 'following columns: '
26 'chrom, chromStart, chromEnd, name, score, strand, '
27 'where "chromStart", "chromEnd" need to be integer values '
28 'and "strand" needs to be either "+" or "-".'
29 )
30
31 # parse amplicon info and record outer primer names
32 with open(sys.argv[2]) as i:
33 ret_lines = []
34 for line in i:
35 first = last = None
36 for pname in line.strip().split('\t'):
37 try:
38 primer_start = primer_starts[pname]
39 except KeyError:
40 sys.exit(
41 'Amplicon info with primer name not found in '
42 f'primer BED file: "{pname}"'
43 )
44 if first is None or primer_start < primer_starts[first]:
45 first = pname
46 if last is None or primer_start > primer_starts[last]:
47 last = pname
48 if first == last:
49 sys.exit(
50 line
51 + 'is not a proper amplicon info line.'
52 )
53 ret_lines.append(f'{first}\t{last}\n')
54
55 # write amended amplicon info
56 with open(sys.argv[3], 'w') as o:
57 o.writelines(ret_lines)