Module: Entrez

Defined in:
lib/rbbt/sources/entrez.rb

Defined Under Namespace

Classes: Gene

Class Method Summary collapse

Class Method Details

.entrez2name(taxs, options = {}) ⇒ Object 



24
25
26
27
28
29
30
31
32
33
34
 
# File 'lib/rbbt/sources/entrez.rb', line 24

def self.entrez2name(taxs, options = {})
  options = Misc.add_defaults options, :key_field => 1, :fields => [2], :persist => true, :merge => true

  taxs = [taxs] unless Array === taxs
  options.merge! :grep => taxs.collect{|t| "^" + t.to_s}
  
  tsv = Rbbt.share.databases.entrez.gene_info.tsv :flat, options
  tsv.key_field = "Entrez Gene ID"
  tsv.fields    = ["Associated Gene Name"]
  tsv
end

.entrez2native(taxs, options = {}) ⇒ Object 



12
13
14
15
16
17
18
19
20
21
22
 
# File 'lib/rbbt/sources/entrez.rb', line 12

def self.entrez2native(taxs, options = {})
  options = Misc.add_defaults options, :key_field => 1, :fields => [5], :persist => true, :merge => true

  taxs = [taxs] unless Array === taxs
  options.merge! :grep => taxs.collect{|t| "^" + t.to_s}
  
  tsv = Rbbt.share.databases.entrez.gene_info.tsv :flat, options
  tsv.key_field = "Entrez Gene ID"
  tsv.fields    = ["Native ID"]
  tsv
end

.entrez2pubmed(taxs) ⇒ Object 



37
38
39
40
41
42
43
44
 
# File 'lib/rbbt/sources/entrez.rb', line 37

def self.entrez2pubmed(taxs)
  options = {:key_field => 1, :fields => [2], :persist => true, :merge => true}

  taxs = [taxs] unless taxs.is_a?(Array)
  options.merge! :grep => taxs.collect{|t| "^" + t.to_s}

  Rbbt.share.databases.entrez.gene2pubmed.tsv :flat, options
end

.gene_filename(id) ⇒ Object 



108
109
110
 
# File 'lib/rbbt/sources/entrez.rb', line 108

def self.gene_filename(id)
  'gene-' + id.to_s + '.xml'
end

.gene_text_similarity(gene, text) ⇒ Object

Counts the words in common between a chunk of text and the text found in Entrez Gene for that particular gene. The gene may be a gene identifier or a Gene class instance.



158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
 
# File 'lib/rbbt/sources/entrez.rb', line 158

def self.gene_text_similarity(gene, text)

  case
  when Entrez::Gene === gene
    gene_text = gene.text
  when String === gene || Fixnum === gene
    begin
      gene_text =  get_gene(gene).text
    rescue CMD::CMDError
      return 0
    end
  else
    return 0
  end

  gene_words = gene_text.words.to_set
  text_words = text.words.to_set

  return 0 if gene_words.empty? || text_words.empty?

  common = gene_words.intersection(text_words)
  common.length / (gene_words.length + text_words.length).to_f
end

.get_gene(geneid) ⇒ Object 



112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
 
# File 'lib/rbbt/sources/entrez.rb', line 112

def self.get_gene(geneid)
  return nil if geneid.nil?

  if Array === geneid
    missing = []
    list = {}

    geneid.each{|p|
      next if p.nil?
      if FileCache.found(gene_filename p)
        list[p] = Gene.new(Open.read(FileCache.path(gene_filename p)))
      else
        missing << p 
      end
    }


    return list unless missing.any?
    genes = get_online(missing)

    genes.each{|p, xml|
      filename = gene_filename p    
      FileCache.add(filename,xml) unless FileCache.found(filename)
      list[p] =  Gene.new(xml)
    }

    return list
  else
    filename = gene_filename geneid    


    if FileCache.found(filename)
      return Gene.new(Open.read(FileCache.path(filename)))
    else
      xml = get_online(geneid)

      FileCache.add(filename, xml) unless FileCache.found(filename)

      return Gene.new(xml)
    end
  end
end