Mercurial > repos > cpt > cpt_get_orfs

comparison get_orfs_or_cdss.py @ 4:f97bc7f587a1 draft

Find changesets by keywords (author, files, the commit message), revision number or hash, or revset expression.
planemo upload commit 94b0cd1fff0826c6db3e7dc0c91c0c5a8be8bb0c
author cpt
date Mon, 05 Jun 2023 02:43:41 +0000
parents
children
comparison
equal deleted inserted replaced
3:29242397252f 4:f97bc7f587a1
1 #!/usr/bin/env python
2 import argparse
3 import logging
4 from cpt import OrfFinder
5
6 logging.basicConfig()
7 log = logging.getLogger()
8
9
10 if __name__ == "__main__":
11 parser = argparse.ArgumentParser(description="Get open reading frames")
12 parser.add_argument("fasta_file", type=argparse.FileType("r"), help="Fasta file")
13
14 parser.add_argument(
15 "-f",
16 "--format",
17 dest="seq_format",
18 default="fasta",
19 help="Sequence format (e.g. fasta, fastq, sff)",
20 )
21 parser.add_argument(
22 "--table", dest="table", default=1, help="NCBI Translation table", type=int
23 )
24 parser.add_argument(
25 "-t",
26 "--ftype",
27 dest="ftype",
28 choices=("CDS", "ORF"),
29 default="ORF",
30 help="Find ORF or CDSs",
31 )
32 parser.add_argument(
33 "-e",
34 "--ends",
35 dest="ends",
36 choices=("open", "closed"),
37 default="closed",
38 help="Open or closed. Closed ensures start/stop codons are present",
39 )
40 parser.add_argument(
41 "-m",
42 "--mode",
43 dest="mode",
44 choices=("all", "top", "one"),
45 default="all",
46 help="Output all ORFs/CDSs from sequence, all ORFs/CDSs "
47 "with max length, or first with maximum length",
48 )
49 parser.add_argument(
50 "--min_len", dest="min_len", default=10, help="Minimum ORF/CDS length", type=int
51 )
52 parser.add_argument(
53 "--strand",
54 dest="strand",
55 choices=("both", "forward", "reverse"),
56 default="both",
57 help="select strand",
58 )
59
60 parser.add_argument(
61 "--on",
62 dest="out_nuc",
63 type=argparse.FileType("w"),
64 default="out.fna",
65 help="Output nucleotide sequences",
66 )
67 parser.add_argument(
68 "--op",
69 dest="out_prot",
70 type=argparse.FileType("w"),
71 default="out.fa",
72 help="Output protein sequences",
73 )
74 parser.add_argument(
75 "--ob",
76 dest="out_bed",
77 type=argparse.FileType("w"),
78 default="out.bed",
79 help="Output BED file",
80 )
81 parser.add_argument(
82 "--og",
83 dest="out_gff3",
84 type=argparse.FileType("w"),
85 default="out.gff3",
86 help="Output GFF3 file",
87 )
88 parser.add_argument("-v", action="version", version="0.3.0")
89 args = parser.parse_args()
90
91 of = OrfFinder(args.table, args.ftype, args.ends, args.min_len, args.strand)
92 of.locate(args.fasta_file, args.out_nuc, args.out_prot, args.out_bed, args.out_gff3)