Class: ChemScanner::PerkinEln

Inherits:
Object
  • Object
show all
Defined in:
lib/chem_scanner/perkin_eln.rb

Overview

Read and Parse ChemDraw ELN XML

Instance Attribute Summary collapse

Instance Method Summary collapse

Constructor Details

#initialize ⇒ PerkinEln

Returns a new instance of PerkinEln.



8
9
10
11
12
13
14
15
16
 
# File 'lib/chem_scanner/perkin_eln.rb', line 8

def initialize
  super

  @molecules = []
  @reactions = []
  @scheme_list = []

  @resolved_ids = []
end

Instance Attribute Details

#molecules ⇒ Object

Returns the value of attribute molecules.



6
7
8
 
# File 'lib/chem_scanner/perkin_eln.rb', line 6

def molecules
  @molecules
end

#reactions ⇒ Object

Returns the value of attribute reactions.



6
7
8
 
# File 'lib/chem_scanner/perkin_eln.rb', line 6

def reactions
  @reactions
end

#scheme_list ⇒ Object

Returns the value of attribute scheme_list.



6
7
8
 
# File 'lib/chem_scanner/perkin_eln.rb', line 6

def scheme_list
  @scheme_list
end

#version ⇒ Object

Returns the value of attribute version.



6
7
8
 
# File 'lib/chem_scanner/perkin_eln.rb', line 6

def version
  @version
end

Instance Method Details

#add_details(key, details) ⇒ Object 



177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
 
# File 'lib/chem_scanner/perkin_eln.rb', line 177

def add_details(key, details)
  return if details.nil? || details.empty?

  if details.class == Array
    unresolved = []

    details.each do |detail|
      mol = try_add_molecule_details(key, detail)
      unresolved.push(detail) if mol.nil?
    end

    unresolved.each do |detail|
      mol = try_match_smi_with_mol(detail)
      add_reaction_details(key, detail) if mol.nil?
    end
  else
    add_reaction_details(key, details)
  end
end

#add_reaction_details(group, detail) ⇒ Object 



273
274
275
276
277
278
279
 
# File 'lib/chem_scanner/perkin_eln.rb', line 273

def add_reaction_details(group, detail)
  group = "Reaction Description" if %w[Reactants Products].include?(group)

  reaction = @reactions.last
  reaction.details[group] = [] if reaction.details[group].nil?
  reaction.details[group].push(detail)
end

#do_cdxml(cdxml) ⇒ Object 



139
140
141
142
143
144
 
# File 'lib/chem_scanner/perkin_eln.rb', line 139

def do_cdxml(cdxml)
  cp = Cdxml.new
  cp.read(cdxml, false)

  cp
end

#do_object(object) ⇒ Object 



53
54
55
56
57
58
59
60
61
62
63
64
65
 
# File 'lib/chem_scanner/perkin_eln.rb', line 53

def do_object(object)
  obj_type = object.at_xpath("field")["name"]
  details = do_object_detail(object)

  obj_details = case obj_type
                when "Preparation" then details["styledText"]
                when "Reaction Conditions" then details["propertyInstances"]
                when "Reaction" then details["chemicalStructure"]
                else details["tableSection"]
                end

  { obj_type => obj_details }
end

#do_object_detail(object) ⇒ Object 



67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
 
# File 'lib/chem_scanner/perkin_eln.rb', line 67

def do_object_detail(object)
  details = {}
  object.element_children.each do |child|
    cname = child.name
    val = case cname
          when "styledText" then child.at_xpath("./text").text
          when "chemicalStructure" then do_cdxml(child.content)
          when "propertyInstances" then do_property_instances(child)
          when "tableSection" then do_table_section(child)
          end

    details[cname] = val unless val.nil?
  end

  details
end

#do_property_instances(prop_instances) ⇒ Object 



84
85
86
87
88
89
90
91
92
93
94
 
# File 'lib/chem_scanner/perkin_eln.rb', line 84

def do_property_instances(prop_instances)
  props = {}

  prop_instances.xpath("./propertyInstance").each do |prop|
    pname = prop.at_xpath("property")["name"]
    val = read_value(prop["minValue"], prop["maxValue"], prop["value"])
    props[pname] = val unless val.nil?
  end

  props
end

#do_section(section) ⇒ Object 



33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
 
# File 'lib/chem_scanner/perkin_eln.rb', line 33

def do_section(section)
  section_info = {}
  section_type = section.at_xpath("sectionType")["name"]

  section.xpath("./object").each do |child|
    obj_info = do_object(child)

    if obj_info.key?("Preparation")
      prep = obj_info.delete("Preparation")
      key = "Reaction Conditions"
      section_info[key] = {} unless section_info.key?(key)
      section_info[key]["Preparation"] = prep
    else
      section_info.merge!(obj_info)
    end
  end

  { "type" => section_type, "details" => section_info }
end

#do_table_props(cols) ⇒ Object 



115
116
117
118
119
120
121
122
123
 
# File 'lib/chem_scanner/perkin_eln.rb', line 115

def do_table_props(cols)
  props = []
  cols.each do |col|
    pname = col.at_xpath("property")["name"]
    props.push(pname)
  end

  props
end

#do_table_rows(rows) ⇒ Object 



125
126
127
128
129
130
131
132
133
134
135
136
137
 
# File 'lib/chem_scanner/perkin_eln.rb', line 125

def do_table_rows(rows)
  rows_values = []
  rows.each do |row|
    tags = row.at_xpath("./tags")
    tags = { "ID" => nil, "parentID" => nil } if tags.nil?
    values = row.xpath("./tableCell").map { |x|
      read_value(x["minValue"], x["maxValue"], x["value"])
    }
    rows_values.push([tags["ID"], tags["parentID"]].concat(values))
  end

  rows_values
end

#do_table_section(table) ⇒ Object 



96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
 
# File 'lib/chem_scanner/perkin_eln.rb', line 96

def do_table_section(table)
  infos = []

  tprops = do_table_props(table.xpath("./tableProperty"))
  props = %w[ID parentID].concat(tprops)
  values = do_table_rows(table.xpath("./tableRow"))

  values.each do |rvalue|
    info = {}
    rvalue.each_with_index do |val, idx|
      info[props[idx]] = val
    end

    infos.push(info)
  end

  infos
end

#read(file, is_path = true) ⇒ Object 



18
19
20
21
22
23
24
25
26
27
28
29
30
31
 
# File 'lib/chem_scanner/perkin_eln.rb', line 18

def read(file, is_path = true)
  fs = is_path ? File.open(file) : file
  xml = Nokogiri::XML(fs)

  infos = []
  sections = xml.xpath("//section")
  sections.each do |section|
    info = do_section(section)
    infos.push(info)
  end

  refine_data(infos)
  true
end

#read_value(min, max, val) ⇒ Object 



146
147
148
149
150
151
152
 
# File 'lib/chem_scanner/perkin_eln.rb', line 146

def read_value(min, max, val)
  mm = min.nil? && max.nil?
  return nil if mm && val.nil?
  return val if mm

  min == max ? val : `${min} ~ ${max}`
end

#refine_data(infos) ⇒ Object 



154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
 
# File 'lib/chem_scanner/perkin_eln.rb', line 154

def refine_data(infos)
  infos.each do |section|
    section_details = section["details"]
    cp = section_details["Reaction"]

    oscheme = OpenStruct.new(
      cdxml: cp,
      molecules: cp.molecules,
      reactions: cp.reactions,
    )
    @scheme_list.push(oscheme)

    @molecules.concat(cp.molecules)
    @reactions.concat(cp.reactions)

    section_details.each do |key, details|
      next if key == "Reaction"

      add_details(key, details)
    end
  end
end

#to_cml(molecule_only = false) ⇒ Object 



281
282
283
284
285
 
# File 'lib/chem_scanner/perkin_eln.rb', line 281

def to_cml(molecule_only = false)
  objs = molecule_only ? @molecules : @reactions
  cml = ChemScanner::Export::CML.new(objs, molecule_only)
  cml.process
end

#try_add_molecule_details(key, detail) ⇒ Object 



197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
 
# File 'lib/chem_scanner/perkin_eln.rb', line 197

def try_add_molecule_details(key, detail)
  id_str = detail["ID"]
  if id_str.nil?
    if key == "Solvents"
      return try_add_solvent(detail)
    else
      return add_reaction_details(key, detail)
    end
  end

  id = id_str.to_i
  mol = @molecules.detect { |m| m.id == id }
  return nil if mol.nil?

  @resolved_ids.push(id)

  detail.delete("Chemical Structure")
  mol.details = detail
end

#try_add_solvent(detail) ⇒ Object 



217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
 
# File 'lib/chem_scanner/perkin_eln.rb', line 217

def try_add_solvent(detail)
  name = detail["Name"]
  return if name.nil? || name.empty?

  smiles = ChemScanner.get_abbreviation(name)
  return if smiles.empty?

  rw_mol = RDKitChem::RWMol.mol_from_smiles(smiles)
  mdl = rw_mol.mol_to_mol_block(true, -1, false)

  mol = OpenStruct.new(
    text: "",
    label: "",
    mdl: mdl,
    cano_smiles: smiles,
    details: detail.merge("Solvent": true),
  )
  @molecules.push(mol)
  @reactions.first.reagents.push(mol)
end

#try_match_smi_with_mol(detail) ⇒ Object 



238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
 
# File 'lib/chem_scanner/perkin_eln.rb', line 238

def try_match_smi_with_mol(detail)
  return nil if !detail.key?("Name") || detail["Name"].empty?

  name = detail["Name"].gsub(/\[.*\]/, "").strip
  return nil if name.nil? || name.empty?

  smiles = ChemScanner.get_abbreviation(name)
  return if smiles.empty?

  rw_mol = RDKitChem::RWMol.mol_from_smiles(smiles)
  cano_smiles = rw_mol.mol_to_smiles(true)
  smiles_arr = cano_smiles.split(".")

  fragments = []
  @molecules.reject { |m| @resolved_ids.include?(m.id) }.each do |m|
    next unless smiles_arr.include?(m.cano_smiles)

    fragments.push(m)
  end
  return nil if fragments.empty? || fragments.size != smiles_arr.size

  mol = fragments.first
  fragments.each do |frag|
    next if frag.id == mol.id

    @molecules.delete_if { |m| frag.id == m.id }
    @reactions.each { |reaction| reaction.delete_molecule_by_id(frag.id) }

    mol.add(frag)
  end
  mol.update_output_formats
  mol.details = detail
  @molecules.push(mol)
end