rbbt-sources 3.0.1 → 3.0.2

checksums.yaml

@@ -1,7 +1,7 @@
 ---
 SHA1:
-  metadata.gz: 4b76cc735fd8fd3ff8f93d84517a8c628fc6c8ac
-  data.tar.gz: 1abc7bb9a8414311f014eed4f1eacfabacf5559f
+  metadata.gz: d0ddae660ee5188e8e9ed08bc38a85a01a3fcbf0
+  data.tar.gz: 42e2f12d00c5147ac2eec9cb087f6094342189c3
 SHA512:
-  metadata.gz: d511a3b4aa38e9b090e57fdbc10888c41a0b98c727993e099ed0bab590fb3375735e57d3ba2744ac32cb186b741236478922b7fa7eb4506122ff70ef106ced62
-  data.tar.gz: f5e962266840d656270a1efe5be92edc6cd56a856cef88909d55f523e258e598c285f85574ed652128111459d4a36d13c3345b4aec3c51eafdef0877164d8a9a
+  metadata.gz: 43b3c30586f24975d6295500dbc447b7c8acdb5aa04361c64bbc6eba6628739810f45dd1360c80b78ab6c479751f7eb182a891e57173661080d2b6aea0034f9c
+  data.tar.gz: 3ee58c4b060418bee1b859bd1387ef97217b14b5d2ef26b0594d3813dee14c64ee983349beb2602b3c293223a242a0da25f45cd90fdf56268ee87df35a426083

data/lib/rbbt/sources/organism.rb

@@ -163,4 +163,44 @@ module Organism
 
     raise "No organism identified for taxid #{taxid}. Supported organism are: #{all_organisms * ", "}"
   end
+
+  def self.gene_list_bases(genes, organism = nil)
+    if genes.respond_to? :orgnanism
+      organism = genes.organism if organism.nil?
+      genes = genes.clean_annotations
+    end
+
+    organism ||= "Hsa"
+
+    @@gene_start_end ||= {}
+    gene_start_end = @@gene_start_end[organism] ||= Organism.gene_positions(organism).tsv(:persist => true, :key_field => "Ensembl Gene ID", :fields => ["Gene Start", "Gene End"], :type => :list, :cast => :to_i, :unmamed => true)
+
+    ranges = genes.collect{|gene| start, eend = gene_start_end[gene]; (start..eend) }
+    Misc.total_length(ranges)
+  end
+
+  def self.gene_list_exon_bases(genes, organism = nil)
+    if genes.respond_to? :orgnanism
+      organism = genes.organism if organism.nil?
+      genes = genes.clean_annotations
+    end
+
+    organism ||= "Hsa"
+
+    @@gene_exons_tsv ||= {}
+    gene_exons = @@gene_exons_tsv[organism] ||= Organism.exons(organism).tsv(:persist => true, :key_field => "Ensembl Gene ID", :fields => ["Ensembl Exon ID"], :type => :flat, :merge => true, :unnamed => true)
+
+    @@exon_range_tsv ||= {}
+    exon_ranges = @@exon_range_tsv[organism] ||= Organism.exons(organism).tsv(:persist => true, :fields => ["Exon Chr Start", "Exon Chr End"], :type => :list, :cast => :to_i, :unnamed => true)
+
+    exons = gene_exons.values_at(*genes).compact.flatten.uniq
+
+    exon_ranges = exons.collect{|exon|
+      Log.low "Exon #{ exon } does not have range" and next if not exon_ranges.include? exon
+      pos = exon_ranges[exon]
+      (pos.first..pos.last)
+    }.compact
+    
+    Misc.total_length(exon_ranges)
+  end
 end

metadata

@@ -1,14 +1,14 @@
 --- !ruby/object:Gem::Specification
 name: rbbt-sources
 version: !ruby/object:Gem::Version
-  version: 3.0.1
+  version: 3.0.2
 platform: ruby
 authors:
 - Miguel Vazquez
 autorequire: 
 bindir: bin
 cert_chain: []
-date: 2014-03-19 00:00:00.000000000 Z
+date: 2014-03-25 00:00:00.000000000 Z
 dependencies:
 - !ruby/object:Gem::Dependency
   name: rbbt-util
@@ -170,7 +170,7 @@ required_rubygems_version: !ruby/object:Gem::Requirement
       version: '0'
 requirements: []
 rubyforge_project: 
-rubygems_version: 2.2.1
+rubygems_version: 2.2.2
 signing_key: 
 specification_version: 4
 summary: Data sources for the Ruby Bioinformatics Toolkit (rbbt)