Merge pull request #5 from MattOates/feature/go_terms

Feature/go terms

Author: MattOates

.gitignore

@@ -29,3 +29,4 @@ nytprof.out
 pm_to_blib
 pod2htm*.tmp
 fatlib/
+.idea/

.travis.yml

@@ -1,7 +1,6 @@
 language: perl
 perl:
-  - "5.24"
-  - "5.22"
+  - "5.36"
   - "5.20"
 
 sudo: false

CHANGES.md

@@ -1,6 +1,6 @@
 ###2019-05-17
    * Updated timeout so that downloads from ELM.eu.org are more likely to succeed.
-   * Cut a new maintenance release 1.4.2
+   * Cut a new maintenance release 1.4.3
 
 ###2016-07-05
     * Majority of refactor complete

README.md

@@ -1,7 +1,7 @@
-mELM [![GitHub version](https://badge.fury.io/gh/MattOates%2Fmelm.svg)](https://badge.fury.io/gh/MattOates%2Fmelm)[![Build Status](https://travis-ci.org/MattOates/melm.svg?branch=master)](https://travis-ci.org/MattOates/melm)[![Coverage Status](https://coveralls.io/repos/github/MattOates/melm/badge.svg?branch=master)](https://coveralls.io/github/MattOates/melm?branch=master)
+mELM [![GitHub version](https://badge.fury.io/gh/MattOates%2Fmelm.svg)](https://badge.fury.io/gh/MattOates%2Fmelm)[![Build Status](https://travis-ci.com/MattOates/melm.svg?branch=master)](https://travis-ci.com/MattOates/melm)[![Coverage Status](https://coveralls.io/repos/github/MattOates/melm/badge.svg?branch=master)](https://coveralls.io/github/MattOates/melm?branch=master)
 ==================
 
-[⬇️ Download latest release v1.4.2](https://github.com/MattOates/melm/releases/download/v1.4.2/melm)
+[⬇️ Download latest release v1.4.3](https://github.com/MattOates/melm/releases/download/v1.4.3/melm)
 
 mELM is a tool for masking or assigning Eukaryotic Linear Motifs to protein sequences. Both TSV/GFF3 output or FASTA is possible. 
 Essentially the tool is a CLI to the ELM.eu.org online resource with additional tools for dealing with short motif assignment within disordered regions.
@@ -32,7 +32,7 @@ If you have used mELM with ANCHOR predictions please cite the following:
 Usage
 =====
 
-        melm [h,v,u,U,c,i,a,G,E,X,m,n,C,P,l,t,M,D,d] <SEQ FILES...>
+        melm [h,v,u,U,c,i,a,G,E,X,m,n,C,P,l,t,M,D,d,g] <SEQ FILES...>
            -h, --help
                this message
            -v, --verbose
@@ -71,6 +71,10 @@ Usage
                if ANCHOR is installed use that to filter ELM output based on ANCHOR's IUPred disorder prediction, only include ELMs that fall within disordered regions
            -d <>, --anchor-datapath=<>
                provide the location of the anchor data path, mELM will otherwise assume it's in the same directory as your anchor binary
+           -g <>, --go-filter=<>
+               turn on GO filtering, only show results for ELM classes that have been associated with the GO ID specified
+           -o <>, --organism-filter=<>
+               turn on organism filtering, only show results for ELM instances that have been observed in a given organism
 
 Example Use Cases
 =================
@@ -91,6 +95,11 @@ Get a GFF3 file for a whole genomes worth of protein annotations
 
            melm --assign --GFF3 human_proteins.fa > human_motifs.gff3
 
+Get another GFF3 file but this time be strict on assignment to those active in native disordered state from the nucleus
+using GO filtering for GO ID 0005634 "nucleus"
+
+           melm --assign --GFF3 --logic-filter --disorder-filter --go-filter=0005634 --organism-filter=sapiens human_proteins.fa > disordered_hiqual_human_nucleus_motifs.gff3
+
 Get the latest ELM classes library for use in another script or by yourself
 
            melm --update --list-classes
@@ -103,7 +112,7 @@ License
 =======
 
            melm - Mask and assign ELM motifs in protein sequence libraries
-           (C) 2014-2019  Dr Matt E. Oates
+           (C) 2014-2022  Dr Matt E. Oates
 
            This program is free software: you can redistribute it and/or modify
            it under the terms of the GNU Affero General Public License as

bin/melm

@@ -1,4 +1,5 @@
 #!/usr/bin/env perl
+use utf8;
 =head1 NAME
 
 I<melm> - mELM masking of assigned ELM motifs
@@ -20,7 +21,7 @@ I<melm> - mELM masking of assigned ELM motifs
         get TSV output of all the ELM instances held in the melm cache
     -a, --assign
         do not mask sequences, instead output a TSV format of all the ELM assignments made per sequence
-    -G, -GFF3
+    -G, --GFF3
         produce GFF3 output when using --assign, useful if you wish to add this assignment to a genome browser or similar
     -E <>, --max-class-expect=<>
         filter out ELM classes based on their annotated expectation, bigger E means allow for more common motifs
@@ -44,6 +45,10 @@ I<melm> - mELM masking of assigned ELM motifs
         if ANCHOR is installed use that to filter ELM output based on ANCHOR's IUPred disorder prediction, only include ELMs that fall within disordered regions
     -d <>, --anchor-datapath=<>
         provide the location of the anchor data path, mELM will otherwise assume it's in the same directory as your anchor binary
+    -g <>, --go-filter=<>
+        turn on GO filtering, only show results for ELM classes that have been associated with the GO ID specified
+    -o <>, --organism-filter=<>
+        turn on organism filtering, only show results for ELM instances that have been observed in a given organism
 
 =head1 DESCRIPTION
 
@@ -88,9 +93,9 @@ Get a GFF3 file for a whole genome's worth of protein annotations
 
     melm --assign --GFF3 human_proteins.fa > human_motifs.gff3
 
-Get another GFF3 file but this time be strict on assignment to those active in native disordered state
+Get another GFF3 file but this time be strict on assignment to those active in native disordered state from the nucleus
 
-    melm --assign --GFF3 --logic-filter --disorder-filter human_proteins.fa > disordered_hiqual_human_motifs.gff3
+    melm --assign --GFF3 --logic-filter --disorder-filter --go-filter=0005634 --organism-filter=sapiens human_proteins.fa > disordered_hiqual_human_nucleus_motifs.gff3
 
 Get the latest ELM classes library for use in another script or by yourself
 
@@ -107,7 +112,7 @@ B<Matt Oates> - I<[email protected]>
 =head1 LICENSE
 
     melm - Mask and assign ELM motifs in protein sequence libraries
-    (C) 2014-2019  Dr Matt E. Oates
+    (C) 2014-2022  Dr Matt E. Oates
 
     This program is free software: you can redistribute it and/or modify
     it under the terms of the GNU Affero General Public License as
@@ -124,9 +129,12 @@ B<Matt Oates> - I<[email protected]>
 
 =head1 EDIT HISTORY
 
+2022-06-10
+   * Added GO data support and bumped the version for release
+
 2019-05-17
    * Updated timeout so that downloads from ELM.eu.org are more likely to succeed.
-   * Cut a new maintenance release 1.4.2
+   * Cut a new maintenance release 1.4.3
 
 2016-07-05 - Matt Oates
     * Majority of refactor complete
@@ -181,7 +189,7 @@ use ELM;
 use ELM::Utils 'get_www';
 
 #Current version of the script
-our $VERSION = "v1.4.2";
+our $VERSION = "v1.4.3";
 
 #User options
 my $help;
@@ -205,8 +213,10 @@ my $disorder_filter;
 my $type;
 my $gff;
 my $anchor_datapath;
+my $go_filter;
+my $organism_filter;
 
-#Flags used h,v,u,U,c,i,a,G,E,X,m,n,C,P,l,t,M,D,d
+#Flags used h,v,u,U,c,i,a,G,E,X,m,n,C,P,l,t,M,D,d,g
 GetOptions(
     "help|h!" => \$help,
     "verbose|v!" => \$verbose,
@@ -227,14 +237,16 @@ GetOptions(
     "morf-filter|M!" => \$morf_filter,
     "disorder-filter|D!" => \$disorder_filter,
     "anchor-datapath|d=s" => \$anchor_datapath,
+    "go-filter|g=s" => \$go_filter,
+    "organism-filter|g=s" => \$organism_filter,
 ) or die "Fatal Error: Problem parsing command-line ".$!;
 
 my @fasta_files = @ARGV;
 
 #Print out some help if it was asked for or if no arguments were given.
 pod2usage(-exitstatus => 0, -verbose => 2) if $help;
 
-pod2usage(-exitstatus => 0, -verbose => 1, -msg => "mELM version $VERSION by Matt Oates (C) 2014-2019. Please provide some sequence files to mask or assign ELM motifs to.") 
+pod2usage(-exitstatus => 0, -verbose => 1, -msg => "mELM version $VERSION by Matt Oates (C) 2014-2022. Please provide some sequence files to mask or assign ELM motifs to.")
     unless $update or $upgrade or $list_classes or $list_instances or scalar @fasta_files >= 1;
 
 my $elm = ELM->new(
@@ -245,6 +257,8 @@ my $elm = ELM->new(
             morf_filter => $morf_filter,
             disorder_filter => $disorder_filter,
             logic_filter => $logic_filter,
+            go_filter => $go_filter,
+            organism_filter => $organism_filter,
             num_elms_threshold => $num_elms,
             anchor => ELM::Anchor->new(anchor_datapath => $anchor_datapath)
 );
@@ -274,6 +288,15 @@ if ($upgrade) {
     exit;
 }
 
+if ($elm->go_filter and not $elm->library->go_terms_version) {
+    say STDERR "You're trying to use a GO filter with no cached GO data, trying to update to fetch GO data";
+    $elm->library->update();
+    if (not $elm->library->go_terms_version) {
+        say STDERR "Something appears to have failed with fetching GO data. Cannot proceed.";
+        exit;
+    }
+}
+
 #Update the ELM classes file and populate elms
 if ($update or not $elm->library->exists) {
     $elm->library->update();

lib/ELM.pm

@@ -1,12 +1,12 @@
-package ELM v1.4.2;
+package ELM v1.4.3;
 =encoding UTF-8
 =head1 NAME
 
 ELM - Class to do analysis with the ELM regex library
 
 =head1 VERSION
 
-Version v1.4.2
+Version v1.4.3
 
 =cut
 
@@ -29,7 +29,9 @@ use Class::Tiny qw(
     max_elm_probability
     morf_filter
     disorder_filter
-    logic_filter), { 
+    logic_filter
+    go_filter
+    organism_filter), {
     num_elms_threshold => sub { 1 },
 	library => sub { ELM::Library->new() },
 	anchor => sub { ELM::Anchor->new() },
@@ -116,6 +118,12 @@ sub assign($self, $elm_name, $regex, $string, $morf_regions, $dis_regions) {
         if ($self->disorder_filter) {
             next unless any_overlap($start,$end,$dis_regions);
         }
+        if ($self->go_filter and $self->library->go_terms_version) {
+            next unless $self->_go_filter_ok($elm_name);
+        }
+        if ($self->organism_filter) {
+            next unless $self->_organism_filter_ok($elm_name);
+        }
         push @ret, [$elm_name, $start, $end, $seq, $prob, $entropy, $entrorate];
     }
     if (@ret < 1) {
@@ -172,6 +180,17 @@ sub _logic_filter_ok($self, $elm_name, $seq, %opt) {
     return (not exists $filters{$seq})?1:0;
 }
 
+sub _go_filter_ok($self, $elm_name) {
+     my %elms = %{ $self->library->elms };
+     return 0 != scalar grep {$self->go_filter eq $_->{go_id}} @{$elms{$elm_name}{go_terms}};
+}
+
+sub _organism_filter_ok($self, $elm_name) {
+     # At least one true positive (TP) example from a substr match to the organism text must exit
+     my %elms = %{ $self->library->elms };
+     return 0 != scalar grep {$_->{logic} eq 'TP' and index(fc $_->{organism}, fc $self->organism_filter ) != -1} @{$elms{$elm_name}{instances}};
+}
+
 =head1 AUTHOR
 
 Matt Oates, C<< <mattoates at gmail.com> >>
@@ -219,7 +238,7 @@ If you have used mELM with ANCHOR predictions please cite the following:
 
 =head1 LICENSE AND COPYRIGHT
 
-Copyright 2019 Matt Oates.
+Copyright 2022 Matt Oates.
 
     This program is free software: you can redistribute it and/or modify
     it under the terms of the GNU Affero General Public License as

lib/ELM/AminoAcids.pm

@@ -1,4 +1,4 @@
-package ELM::AminoAcids v1.4.2;
+package ELM::AminoAcids v1.4.3;
 require Exporter;
 =encoding UTF-8
 =head1 NAME
@@ -7,7 +7,7 @@ ELM::AminoAcids - Functions for dealing with amino acid specific calculations
 
 =head1 VERSION
 
-Version v1.4.2
+Version v1.4.3
 
 =cut
 
@@ -142,7 +142,7 @@ If you have used mELM with ANCHOR predictions please cite the following:
 
 =head1 LICENSE AND COPYRIGHT
 
-Copyright 2019 Matt Oates.
+Copyright 2022 Matt Oates.
 
     This program is free software: you can redistribute it and/or modify
     it under the terms of the GNU Affero General Public License as

lib/ELM/Anchor.pm

@@ -1,12 +1,12 @@
-package ELM::Anchor v1.4.2;
+package ELM::Anchor v1.4.3;
 =encoding UTF-8
 =head1 NAME
 
 ELM::Anchor - Class to wrap ANCHOR and get assignments
 
 =head1 VERSION
 
-Version v1.4.2
+Version v1.4.3
 
 =cut
 
@@ -44,11 +44,9 @@ To create an ELM::Anchor explicitly.
 Get/Set the ANCHOR datapath, default to install directory
 
 =cut
-sub anchor_datapath($self) {
+sub anchor_datapath($self, $path) {
     my $defaults = Class::Tiny->get_all_attribute_defaults_for( ref $self );
-    if (@_) {
-        my $path = shift;
-        $path = $defaults->{anchor_datapath}->() unless $path;
+    if ($path) {
         return $self->{anchor_datapath} = $path;
     }
     elsif ( exists $self->{anchor_datapath} ) {
@@ -194,7 +192,7 @@ If you have used mELM with ANCHOR predictions please cite the following:
 
 =head1 LICENSE AND COPYRIGHT
 
-Copyright 2019 Matt Oates.
+Copyright 2022 Matt Oates.
 
     This program is free software: you can redistribute it and/or modify
     it under the terms of the GNU Affero General Public License as

lib/ELM/Calc.pm

@@ -1,4 +1,4 @@
-package ELM::Calc v1.4.2;
+package ELM::Calc v1.4.3;
 require Exporter;
 =encoding UTF-8
 =head1 NAME
@@ -7,7 +7,7 @@ ELM::Calc - Functions for calculating sequence assignment specific tasks
 
 =head1 VERSION
 
-Version v1.4.2
+Version v1.4.3
 
 =cut
 
@@ -138,7 +138,7 @@ If you have used mELM with ANCHOR predictions please cite the following:
 
 =head1 LICENSE AND COPYRIGHT
 
-Copyright 2019 Matt Oates.
+Copyright 2022 Matt Oates.
 
     This program is free software: you can redistribute it and/or modify
     it under the terms of the GNU Affero General Public License as