rbbt-sources 2.0.0 → 2.0.1

data/lib/rbbt/sources/InterPro.rb

@@ -7,16 +7,16 @@ module InterPro
   extend Resource
   self.subdir = "share/databases/InterPro"
 
-  InterPro.claim InterPro.source.protein2ipr.find, :url, "ftp://ftp.ebi.ac.uk/pub/databases/interpro/protein2ipr.dat.gz"
+  InterPro.claim InterPro.source.protein2ipr, :url, "ftp://ftp.ebi.ac.uk/pub/databases/interpro/protein2ipr.dat.gz"
 
-  InterPro.claim InterPro.protein_domains.find, :proc do
+  InterPro.claim InterPro.protein_domains, :proc do
     organism = "Hsa"
     uniprot_colum = TSV::Parser.new(Organism.protein_identifiers(organism).open).all_fields.index("UniProt/SwissProt Accession")
     uniprots = CMD.cmd("grep -v  '^#'|cut -f #{uniprot_colum+1}", :in => Organism.protein_identifiers(organism).open).read.split("\n").collect{|l| l.split("|")}.flatten.uniq.reject{|l| l.empty?}
    
     tsv = nil
     TmpFile.with_file(uniprots * "\n") do |tmpfile|
-        tsv = TSV.open(CMD.cmd("cut -f 1,2,5,6 | sort -u |grep -w -f #{ tmpfile }", :in => InterPro.source.protein2ipr.open, :pipe => true), :merge => true, :type => :double)
+        tsv = TSV.open(CMD.cmd("cut -f 1,2,5,6 | sort -u |grep -w -F -f #{ tmpfile }", :in => InterPro.source.protein2ipr.open, :pipe => true), :merge => true, :type => :double)
     end
  
     tsv.key_field = "UniProt/SwissProt Accession"
@@ -24,7 +24,7 @@ module InterPro
     tsv.to_s
   end
 
-  InterPro.claim InterPro.domain_names.find, :proc do
+  InterPro.claim InterPro.domain_names, :proc do
     #tsv = InterPro.source.protein2ipr.tsv :key_field => 1, :fields => [2], :type => :single
     tsv = TSV.open(CMD.cmd("cut -f 2,3 | sort -u", :in => InterPro.source.protein2ipr.open, :pipe => true), :merge => true, :type => :single)
  
@@ -77,7 +77,7 @@ if defined? Entity
 
     property :genes => :array2single do
       InterPro.gene_index.values_at(*self).
-        collect{|genes| genes = genes.uniq;  genes.organism = organism if genes.respond_to? :organism; genes }.tap{|o| Gene.setup(o, "UniProt/SwissProt Accession", organism)}
+        collect{|genes| genes = [] if genes.nil?; genes = genes.uniq;  genes.organism = organism if genes.respond_to? :organism; genes }.tap{|o| Gene.setup(o, "UniProt/SwissProt Accession", organism)}
     end
   end
 

data/lib/rbbt/sources/genomes1000.rb

@@ -8,22 +8,29 @@ module Genomes1000
 
   RELEASE_URL = "ftp://ftp.1000genomes.ebi.ac.uk/vol1/ftp/release/20110521/ALL.wgs.phase1_release_v3.20101123.snps_indels_sv.sites.vcf.gz"
 
-  Genomes1000.claim Genomes1000.mutations, :proc do
+  Genomes1000.claim Genomes1000.mutations, :proc do |filename|
 
-    tsv = TSV.setup({}, :key_field => "Variant ID", :fields => ["Genomic Mutation"], :type => :single)
-    Open.read(RELEASE_URL) do |line|
-      next if line[0] == "#"[0]
+    begin
+      Open.write(filename) do |file|
+        file.puts "#: :type=:single#:namespace=Hsa"
+        file.puts "#Variant ID\tGenomic Mutation"
 
-      chromosome, position, id, references, alternative, quality, filter, info = line.split("\t")
+        Open.read(RELEASE_URL) do |line|
+          next if line[0] == "#"[0]
 
-      tsv[id] = [chromosome, position, alternative] * ":"
-    end
-
-    tsv.namespace = "Hsa"
+          chromosome, position, id, references, alternative, quality, filter, info = line.split("\t")
 
-    tsv.to_s
+          file.puts [id, [chromosome, position, alternative] * ":"] * "\t"
+        end
+      end
+    rescue
+      FileUtils.rm filename if File.exists? filename
+      raise $!
+    end
+    nil
   end
 
+
   Genomes1000.claim Genomes1000.mutations_hg18, :proc do
     require 'rbbt/sources/organism'
 

data/lib/rbbt/sources/pfam.rb

@@ -31,7 +31,7 @@ module InterPro
   end
 end
 
-InterPro.claim InterPro.pfam_names.find, :proc do
+InterPro.claim InterPro.pfam_names, :proc do
   pfam_domains = Pfam.domains.read.split("\n").collect{|l| l.split("\t").first}.compact.flatten
   tsv = nil
   TmpFile.with_file(pfam_domains * "\n") do |tmpfile|
@@ -42,7 +42,7 @@ InterPro.claim InterPro.pfam_names.find, :proc do
   tsv.to_s
 end
 
-InterPro.claim InterPro.pfam_equivalences.find, :proc do
+InterPro.claim InterPro.pfam_equivalences, :proc do
   pfam_domains = Pfam.domains.read.split("\n").collect{|l| l.split("\t").first}.compact.flatten
   tsv = nil
   TmpFile.with_file(pfam_domains * "\n") do |tmpfile|

metadata

@@ -1,7 +1,7 @@
 --- !ruby/object:Gem::Specification
 name: rbbt-sources
 version: !ruby/object:Gem::Version
-  version: 2.0.0
+  version: 2.0.1
   prerelease: 
 platform: ruby
 authors:
@@ -9,7 +9,7 @@ authors:
 autorequire: 
 bindir: bin
 cert_chain: []
-date: 2012-12-21 00:00:00.000000000 Z
+date: 2013-01-07 00:00:00.000000000 Z
 dependencies:
 - !ruby/object:Gem::Dependency
   name: rbbt-util
@@ -136,12 +136,12 @@ files:
 - share/install/Organism/organism_helpers.rb
 - share/install/STITCH/Rakefile
 - share/install/lib/helpers.rb
-- test/test_helper.rb
-- test/rbbt/sources/test_entrez.rb
-- test/rbbt/sources/test_pubmed.rb
 - test/rbbt/sources/test_go.rb
+- test/rbbt/sources/test_entrez.rb
 - test/rbbt/sources/test_biomart.rb
 - test/rbbt/sources/test_organism.rb
+- test/rbbt/sources/test_pubmed.rb
+- test/test_helper.rb
 homepage: http://github.com/mikisvaz/rbbt-sources
 licenses: []
 post_install_message: 
@@ -167,9 +167,9 @@ signing_key:
 specification_version: 3
 summary: Data sources for the Ruby Bioinformatics Toolkit (rbbt)
 test_files:
-- test/test_helper.rb
-- test/rbbt/sources/test_entrez.rb
-- test/rbbt/sources/test_pubmed.rb
 - test/rbbt/sources/test_go.rb
+- test/rbbt/sources/test_entrez.rb
 - test/rbbt/sources/test_biomart.rb
 - test/rbbt/sources/test_organism.rb
+- test/rbbt/sources/test_pubmed.rb
+- test/test_helper.rb