Mercurial > repos > iuc > gff_to_prot

comparison gff_to_prot.py @ 0:99810cf51f2e 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/transit/ commit 73c6b2baf9dda26c6809a4f36582f7cbdb161ea1
author iuc
date Mon, 22 Apr 2019 14:42:21 -0400
parents
children
comparison
equal deleted inserted replaced
-1:000000000000 0:99810cf51f2e
1 #!/usr/bin/env python
2 import csv
3 import os
4 import sys
5
6
7 def get_description(line, parent):
8 cols = line.split('\t')
9 labels = {}
10 for pair in cols[8].split(";"):
11 k, v = pair.split('=')
12 labels[k] = v
13
14 if (cols[2]) == "CDS" and labels["Parent"] == parent:
15 return labels.get("Note", '-')
16 return '-'
17
18
19 def convert_to_prot_table(fname, output_name):
20 gff_file = open(fname)
21 output_file = open(output_name, 'w')
22 writer = csv.writer(output_file, delimiter='\t')
23 lines = gff_file.readlines()
24 gff_file.close()
25 for i, line in enumerate(lines):
26 line = line.strip()
27 if line.startswith('#'):
28 continue
29 cols = line.split('\t')
30 if (len(cols) < 9):
31 print("Ignoring invalid row with entries: {0}".format(cols))
32 elif (cols[2]) == "region":
33 continue
34 elif (cols[2]) == "CDS":
35 continue
36 elif (cols[2]) == "gene":
37 start = int(cols[3])
38 end = int(cols[4])
39 strand = cols[6].strip()
40 labels = {}
41 diff = int(abs(end - start) / 3) # What is this called?
42 for pair in cols[8].split(";"):
43 k, v = pair.split('=')
44 labels[k.strip()] = v.strip()
45
46 Rv = labels["locus_tag"].strip() # error out if not found
47 gene = labels.get('Name', '')
48 desc = get_description(lines[i + 1], labels.get("ID", "")) if (i + 1) < len(lines) else '-'
49 vals = [desc, start, end, strand, diff, '-', '-', gene, Rv, '-']
50 writer.writerow(vals)
51 output_file.close()
52
53
54 if __name__ == "__main__":
55 usage_string = "Usage: python gff-prot-converter.py <gff filename> <output filename>"
56
57 if len(sys.argv) < 3:
58 print(usage_string)
59 sys.exit(0)
60 file_name = sys.argv[1]
61 if not os.path.exists(file_name):
62 print("File not found. Exiting...")
63 print(usage_string)
64 sys.exit(0)
65 convert_to_prot_table(file_name, sys.argv[2])