rbbt-sources 3.0.37 → 3.1.0

checksums.yaml

@@ -1,7 +1,7 @@
 ---
 SHA1:
-  metadata.gz: 27d9b524e74122ec28214154aae2646350075b34
-  data.tar.gz: 5a41416cadec47a6a65b5dc0fd15f836d8aaf27c
+  metadata.gz: b163305cdfa2b814e25b239b7a7d8780a473beba
+  data.tar.gz: 914b6e2b8c118d2d2171a3d648a376ed94521c6d
 SHA512:
-  metadata.gz: f7479d80768eecbd609d4bb15907a36eb142d693ce185122fe7c4fe90d8b66cafa10aa0454dabc22db0a6ef4f73500d60752ef6f317b81c495929bfb7756bd2b
-  data.tar.gz: c9227fd124bad413ca05292702b3adcd9fe10bacbbb242022621c64c9fa63cf14ff218158dc93a48b6805e27872fb06e22f24f9cc00555896a12ef7566999037
+  metadata.gz: 0aa5a1669673cb9a716edab3ee4cecd853482bdf8e39f3663b0041a725b3cbacd74678db5358613734e180aa65ff3a46294d17a124343efc9d92986ec75ec8bb
+  data.tar.gz: 944341b1833b892e059dcc272747981ea5557b295cd9bcb06eb3886f4ad9652a15489532fab32d487b59047e88b18f386200b0190184b1cfe0b63a02c3387872

data/etc/biomart/missing_in_archive

@@ -1,4 +1,4 @@
-all:
+__all:
     - refseq_mrna
 ">jun2015":
     - uniprot_swissprot

data/share/install/Organism/Hsa/Rakefile

@@ -48,6 +48,10 @@ $biomart_probe_identifiers = [
   [ 'Codelink ID', 'codelink' ],
   [ 'Illumina HumanWG 6 v2', 'illumina_humanwg_6_v2' ],
   [ 'Illumina HumanWG 6 v3', 'illumina_humanwg_6_v3' ],
+  [ 'RefSeq mRNA' , "refseq_mrna"] , 
+  [ 'RefSeq ncRNA' , "refseq_ncrna"] , 
+  [ 'RefSeq mRNA predicted' , "refseq_mrna_predicted"] , 
+  [ 'RefSeq ncRNA predicted' , "refseq_ncrna_predicted"] , 
 ]
 
 $biomart_identifiers = [ 
@@ -56,13 +60,16 @@ $biomart_identifiers = [
   [ 'Associated Gene Name', "external_gene_name"  ],
   [ 'CCDS ID', "ccds"  ],
   [ 'Protein ID', "protein_id"  ],
+  [ 'RefSeq mRNA' , "refseq_mrna"] , 
+  [ 'RefSeq ncRNA' , "refseq_ncrna"] , 
+  [ 'RefSeq mRNA predicted' , "refseq_mrna_predicted"] , 
+  [ 'RefSeq ncRNA predicted' , "refseq_ncrna_predicted"] , 
   [ 'RefSeq Protein ID', "refseq_peptide"  ],
   [ 'Unigene ID', "unigene"  ],
   [ 'UniProt/SwissProt ID', "uniprot_swissprot"  ],
   [ 'UniProt/SwissProt Accession', "uniprot_swissprot_accession"  ],
   [ 'HGNC ID', "hgnc_id", 'HGNC'],
   #[ 'EMBL (Genbank) ID' , "embl"] , 
-  #[ 'RefSeq mRNA' , "refseq_mrna"] , 
 
   # Probes
   [ 'AFFY HC G110', 'affy_hc_g110' ],

data/share/install/Organism/organism_helpers.rb

@@ -127,12 +127,27 @@ file 'identifiers' do |t|
     end
   end
 
+  refseq_fields = identifiers.fields.select{|f| f =~ /RefSeq/}
+
+  if refseq_fields.any?
+    refseq_pos = refseq_fields.collect{|refseq_field| identifiers.identify_field refseq_field }
+    identifiers = identifiers.add_field "RefSeq ID" do |key,values|
+      refseq_ids = values.values_at *refseq_pos
+      refseq_ids.flatten.compact.reject{|v| v.empty?}
+    end
+
+    ordered_fields = identifiers.fields - ["RefSeq ID"]
+    pos = ordered_fields.index ordered_fields.select{|f| f =~ /RefSeq/}.first
+    ordered_fields[pos..pos-1] = "RefSeq ID"
+
+    identifiers = identifiers.reorder(:key, ordered_fields)
+  end
+
   entrez_synonyms = Rbbt.share.databases.entrez.gene_info.find.tsv :grep => $taxs.collect{|tax| "^#{tax}"}, :key_field => 1, :fields => [4]
   entrez_synonyms.key_field = "Entrez Gene ID"
   entrez_synonyms.fields = ["Entrez Gene Name Synonyms"]
 
   identifiers.attach entrez_synonyms
-  
 
   identifiers.with_unnamed do
     identifiers.each do |key, values|
@@ -143,6 +158,7 @@ file 'identifiers' do |t|
     end
   end
 
+
   Misc.sensiblewrite(t.name, identifiers.to_s)
 end
 

metadata

@@ -1,14 +1,14 @@
 --- !ruby/object:Gem::Specification
 name: rbbt-sources
 version: !ruby/object:Gem::Version
-  version: 3.0.37
+  version: 3.1.0
 platform: ruby
 authors:
 - Miguel Vazquez
 autorequire: 
 bindir: bin
 cert_chain: []
-date: 2015-11-20 00:00:00.000000000 Z
+date: 2015-11-23 00:00:00.000000000 Z
 dependencies:
 - !ruby/object:Gem::Dependency
   name: rbbt-util