Mercurial > repos > galaxyp > fragpipe

annotate genericize_db.py @ 0:41990c43f371 draft

Find changesets by keywords (author, files, the commit message), revision number or hash, or revset expression.
planemo upload for repository https://github.com/galaxyproteomics/tools-galaxyp/tree/master/tools/fragpipe commit 905cc2be18669cffe9ac6c46fcd08b6857a67f4f
author galaxyp
date Wed, 10 Jul 2024 06:14:45 +0000
parents
children
Ignore whitespace changes - Everywhere: Within whitespace: At end of lines:
rev   line source
0
41990c43f371 planemo upload for repository https://github.com/galaxyproteomics/tools-galaxyp/tree/master/tools/fragpipe commit 905cc2be18669cffe9ac6c46fcd08b6857a67f4f
galaxyp
parents:
diff changeset
1 #!/usr/bin/env python3
41990c43f371 planemo upload for repository https://github.com/galaxyproteomics/tools-galaxyp/tree/master/tools/fragpipe commit 905cc2be18669cffe9ac6c46fcd08b6857a67f4f
galaxyp
parents:
diff changeset
2 #
41990c43f371 planemo upload for repository https://github.com/galaxyproteomics/tools-galaxyp/tree/master/tools/fragpipe commit 905cc2be18669cffe9ac6c46fcd08b6857a67f4f
galaxyp
parents:
diff changeset
3 # Prefixes sequence headers in the input FASTA file that are not formatted according to the UniProt, NCBI, or ENSEMBL formats with '>generic|' to avoid being misinterpreted by Philosopher.
41990c43f371 planemo upload for repository https://github.com/galaxyproteomics/tools-galaxyp/tree/master/tools/fragpipe commit 905cc2be18669cffe9ac6c46fcd08b6857a67f4f
galaxyp
parents:
diff changeset
4 #
41990c43f371 planemo upload for repository https://github.com/galaxyproteomics/tools-galaxyp/tree/master/tools/fragpipe commit 905cc2be18669cffe9ac6c46fcd08b6857a67f4f
galaxyp
parents:
diff changeset
5
41990c43f371 planemo upload for repository https://github.com/galaxyproteomics/tools-galaxyp/tree/master/tools/fragpipe commit 905cc2be18669cffe9ac6c46fcd08b6857a67f4f
galaxyp
parents:
diff changeset
6 import re
41990c43f371 planemo upload for repository https://github.com/galaxyproteomics/tools-galaxyp/tree/master/tools/fragpipe commit 905cc2be18669cffe9ac6c46fcd08b6857a67f4f
galaxyp
parents:
diff changeset
7 import sys
41990c43f371 planemo upload for repository https://github.com/galaxyproteomics/tools-galaxyp/tree/master/tools/fragpipe commit 905cc2be18669cffe9ac6c46fcd08b6857a67f4f
galaxyp
parents:
diff changeset
8
41990c43f371 planemo upload for repository https://github.com/galaxyproteomics/tools-galaxyp/tree/master/tools/fragpipe commit 905cc2be18669cffe9ac6c46fcd08b6857a67f4f
galaxyp
parents:
diff changeset
9 input_db_file = sys.argv[1]
41990c43f371 planemo upload for repository https://github.com/galaxyproteomics/tools-galaxyp/tree/master/tools/fragpipe commit 905cc2be18669cffe9ac6c46fcd08b6857a67f4f
galaxyp
parents:
diff changeset
10 output_db_file = sys.argv[2]
41990c43f371 planemo upload for repository https://github.com/galaxyproteomics/tools-galaxyp/tree/master/tools/fragpipe commit 905cc2be18669cffe9ac6c46fcd08b6857a67f4f
galaxyp
parents:
diff changeset
11
41990c43f371 planemo upload for repository https://github.com/galaxyproteomics/tools-galaxyp/tree/master/tools/fragpipe commit 905cc2be18669cffe9ac6c46fcd08b6857a67f4f
galaxyp
parents:
diff changeset
12
41990c43f371 planemo upload for repository https://github.com/galaxyproteomics/tools-galaxyp/tree/master/tools/fragpipe commit 905cc2be18669cffe9ac6c46fcd08b6857a67f4f
galaxyp
parents:
diff changeset
13 def sub_header(line):
41990c43f371 planemo upload for repository https://github.com/galaxyproteomics/tools-galaxyp/tree/master/tools/fragpipe commit 905cc2be18669cffe9ac6c46fcd08b6857a67f4f
galaxyp
parents:
diff changeset
14 return re.sub(r'^>(?!sp\||tr\||db\||AP_|NP_|YP_|XP_|WP_|ENSP|UniRef|nxp|generic)', '>generic|', line)
41990c43f371 planemo upload for repository https://github.com/galaxyproteomics/tools-galaxyp/tree/master/tools/fragpipe commit 905cc2be18669cffe9ac6c46fcd08b6857a67f4f
galaxyp
parents:
diff changeset
15
41990c43f371 planemo upload for repository https://github.com/galaxyproteomics/tools-galaxyp/tree/master/tools/fragpipe commit 905cc2be18669cffe9ac6c46fcd08b6857a67f4f
galaxyp
parents:
diff changeset
16
41990c43f371 planemo upload for repository https://github.com/galaxyproteomics/tools-galaxyp/tree/master/tools/fragpipe commit 905cc2be18669cffe9ac6c46fcd08b6857a67f4f
galaxyp
parents:
diff changeset
17 with open(input_db_file) as in_file, open(output_db_file, 'w') as out_file:
41990c43f371 planemo upload for repository https://github.com/galaxyproteomics/tools-galaxyp/tree/master/tools/fragpipe commit 905cc2be18669cffe9ac6c46fcd08b6857a67f4f
galaxyp
parents:
diff changeset
18 for line in in_file:
41990c43f371 planemo upload for repository https://github.com/galaxyproteomics/tools-galaxyp/tree/master/tools/fragpipe commit 905cc2be18669cffe9ac6c46fcd08b6857a67f4f
galaxyp
parents:
diff changeset
19 out_file.write(sub_header(line))