Mercurial > repos > cpt > cpt_psm_prep

comparison lib/CPT/Bio/GFF_Parsing.pm @ 1:d724f34e671d draft default tip

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 02:50:07 +0000
parents
children
comparison
equal deleted inserted replaced
0:e4de0a0e90c8 1:d724f34e671d
1 package CPT::Bio::GFF_Parsing;
2 use Moose;
3 use autodie;
4
5 my $tags = 'allele anticodon artificial_location
6 bio_material bound_moiety cell_line cell_type chromosome citation
7 clone clone_lib codon_start collected_by collection_date compare
8 country cultivar culture_collection db_xref dev_stage direction
9 EC_number ecotype environmental_sample estimated_length exception
10 experiment focus frequency function gap_type gene gene_synonym
11 germline haplogroup haplotype host identified_by inference isolate
12 isolation_source lab_host lat_lon linkage_evidence locus_tag
13 macronuclear map mating_type mobile_element_type mod_base mol_type
14 ncRNA_class note number old_locus_tag operon organelle organism
15 partial PCR_conditions PCR_primers phenotype plasmid pop_variant
16 product protein_id proviral pseudo rearranged replace
17 ribosomal_slippage rpt_family rpt_type rpt_unit_range rpt_unit_seq
18 satellite segment serotype serovar sex specimen_voucher
19 standard_name strain sub_clone sub_species sub_strain tag_peptide
20 tissue_lib tissue_type transgenic translation transl_except
21 transl_table trans_splicing variety';
22 my %valid_tags = map { $_ => 1 } split( /\s+/, $tags );
23
24 my $keys = "-10_signal -35_signal 3'UTR 5'UTR
25 CAAT_signal CDS C_region D-loop D_segment GC_signal J_segment LTR
26 N_region RBS STS S_region TATA_signal V_region V_segment
27 assembly_gap attenuator enhancer exon gap gene iDNA intron mRNA
28 mat_peptide misc_RNA misc_binding misc_difference misc_feature
29 misc_recomb misc_signal misc_structure mobile_element
30 modified_base ncRNA old_sequence operon oriT polyA_signal
31 polyA_site precursor_RNA prim_transcript primer_bind promoter
32 protein_bind rRNA rep_origin repeat_region sig_peptide source
33 stem_loop tRNA terminator tmRNA transit_peptide unsure variation";
34 my %valid_keys = map { $_ => 1 } split( /\s+/, $keys );
35
36 has 'tag_conv' => ( is => 'ro', isa => 'HashRef', default => sub { return {
37 #tags
38 'dbxref' => 'db_xref',
39 'parent' => 'note',
40 'name' => 'label',
41 'Name' => 'label',
42 'old-name' => 'obsolete_name',
43 'id' => 'note',
44 'nat-host' => 'host',
45 'genome' => 'note',
46 'region' => 'source',
47 }; });
48
49
50 sub fix_gff_tag {
51 my ($self, $tag) = @_;
52 # Lowercase it
53 if(defined $tag && defined(${$self->tag_conv()}{lc($tag)})){
54 return ${$self->tag_conv()}{lc($tag)};
55 }
56 return $tag;
57 }
58
59
60 no Moose;
61 1;
62
63 __END__
64
65 =pod
66
67 =encoding UTF-8
68
69 =head1 NAME
70
71 CPT::Bio::GFF_Parsing
72
73 =head1 VERSION
74
75 version 1.99.4
76
77 =head1 AUTHOR
78
79 Eric Rasche <rasche.eric@yandex.ru>
80
81 =head1 COPYRIGHT AND LICENSE
82
83 This software is Copyright (c) 2014 by Eric Rasche.
84
85 This is free software, licensed under:
86
87 The GNU General Public License, Version 3, June 2007
88
89 =cut