Module: UniProt

Extended by:
Resource
Defined in:
lib/rbbt/sources/uniprot.rb

Constant Summary collapse

UNIPROT_TEXT = 
"http://www.uniprot.org/uniprot/[PROTEIN].txt"
UNIPROT_FASTA = 
"http://www.uniprot.org/uniprot/[PROTEIN].fasta"

Class Method Summary collapse

Class Method Details

.cath(protein) ⇒ Object 



159
160
161
162
163
164
165
166
167
168
169
170
 
# File 'lib/rbbt/sources/uniprot.rb', line 159

def self.cath(protein)
  url = UNIPROT_TEXT.sub "[PROTEIN]", protein
  text = Open.read(url)

  cath = {}
  text.split(/\n/).each{|l| 
    next unless l =~ /^DR\s+Gene3D; G3DSA:(.*)\./
    id, description, cuantity = $1.split(";").collect{|v| v.strip}
    cath[id] = {:description => description, :cuantity => cuantity}
  }
  cath
end

.cath_domains(protein) ⇒ Object 



172
173
174
175
176
177
 
# File 'lib/rbbt/sources/uniprot.rb', line 172

def self.cath_domains(protein)
  pdbs = pdbs(protein).keys.uniq
  pdbs.collect do |pdb|
    Cath.domains_for_pdb(pdb)
  end.flatten.compact
end

.features(protein) ⇒ Object 



68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
 
# File 'lib/rbbt/sources/uniprot.rb', line 68

def self.features(protein)
  url = UNIPROT_TEXT.sub "[PROTEIN]", protein
  text = Open.read(url)

  text = text.split(/\n/).select{|line| line =~ /^FT/} * "\n"

  parts = text.split(/^(FT   \w+)/)
  parts.shift

  features = []

  type = nil
  parts.each do |part|
    parts
    if part[0..1] == "FT"
      type = part.gsub(/FT\s+/,'')
      next
    end
    value = part.gsub("\nFT", '').gsub(/\s+/, ' ')
    case
    when value.match(/(\d+) (\d+) (.*)/)
      start, eend, description = $1, $2, $3
      description.gsub(/^FT\s+/m, '')
    when value.match(/(\d+) (\d+)/)
      start, eend = $1, $2
      description = nil
    else
      Log.debug "Value not understood: #{ value }"
    end


    feature = {
      :type => type,
      :start => start.to_i, 
      :end => eend.to_i, 
      :description => description,
    }

    features << feature
  end

  features
end

.pdbs(protein) ⇒ Object 



38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
 
# File 'lib/rbbt/sources/uniprot.rb', line 38

def self.pdbs(protein)
  url = UNIPROT_TEXT.sub "[PROTEIN]", protein
  text = Open.read(url)

  pdb = {}

  text.split(/\n/).each{|l| 
    next unless l =~ /^DR\s+PDB; (.*)\./
    id, method, resolution, region = $1.split(";").collect{|v| v.strip}
    begin
      chains, start, eend = region.match(/(\w+)=(\d+)-(\d+)/).values_at(1,2,3)
      start = start.to_i
      eend = eend.to_i
      start, eend = eend, start if start > eend
    rescue
      Log.warn("Error process Uniprot PDB line: #{line}")
      next
    end
    pdb[id.downcase] = {:method => method, :resolution => resolution, :region => (start..eend), :chains => chains}
  }
  pdb
end

.pdbs_covering_aa_position(protein, aa_position) ⇒ Object 



179
180
181
182
183
 
# File 'lib/rbbt/sources/uniprot.rb', line 179

def self.pdbs_covering_aa_position(protein, aa_position)
  UniProt.pdbs(protein).select do |pdb, info|
    info[:region].include? aa_position
  end
end

.sequence(protein) ⇒ Object 



61
62
63
64
65
66
 
# File 'lib/rbbt/sources/uniprot.rb', line 61

def self.sequence(protein)
  url = UNIPROT_FASTA.sub "[PROTEIN]", protein
  text = Open.read(url)

  text.split(/\n/).select{|line| line !~ /^>/} * ""
end

.variants(protein) ⇒ Object 



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
154
155
156
157
 
# File 'lib/rbbt/sources/uniprot.rb', line 113

def self.variants(protein)
  url = UNIPROT_TEXT.sub "[PROTEIN]", protein
  text = Open.read(url)

  text = text.split(/\n/).select{|line| line =~ /^FT/} * "\n"

  parts = text.split(/^(FT   \w+)/)
  parts.shift

  variants = []

  type = nil
  parts.each do |part|
    if type.nil?
      type = part
    else
      if type !~ /VARIANT/
        type = nil
        next
      end
      type = nil
      
      value = part.gsub("\nFT", '').gsub(/\s+/, ' ')
      # 291 291 K -> E (in sporadic cancers; somatic mutation). /FTId=VAR_045413.
      case
      when value.match(/(\d+) (\d+) ([A-Z])\s*\-\>\s*([A-Z]) (.*)\. \/FTId=(.*)/)
        start, eend, ref, mut, desc, id = $1, $2, $3, $4, $5, $6
      when value.match(/(\d+) (\d+) (.*)\. \/FTId=(.*)/)
        start, eend, ref, mut, desc, id = $1, $2, nil, nil, $3, $4
      else
        Log.debug "Value not understood: #{ value }"
      end
      variants << {
        :start => start, 
        :end => eend, 
        :ref => ref,
        :mut => mut, 
        :desc => desc, 
        :id => id,
      }
    end
  end

  variants
end