1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
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
111
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
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
|
require 'rubygems'
require 'compass' #must be loaded before sinatra
require 'sinatra'
require 'sinatra/contrib'
require 'haml' #must be loaded after sinatra
require 'opentox-client'
require 'opentox-server'
require_relative 'helper.rb'
require File.join(ENV["HOME"],".opentox","config","lazar-gui.rb") # until added to ot-tools
# DG: workaround for https://github.com/sinatra/sinatra/issues/808
# Date: 18/11/2013
set :protection, :except => :path_traversal
helpers do
# get prediction models from text file, ignore validation models
# model uris must be manually added
@@models = []
CSV.foreach("./prediction_models.csv"){|uri| m = OpenTox::Model::Lazar.find uri[0]; @@models << m}
end
get '/?' do
redirect to('/predict')
end
get '/predict/?' do
# sort models by endpoint alphabetically
@models = @@models.sort!{|a, b| a.type.select{|e| e =~ /endpoint/i} <=> b.type.select{|e| e =~ /endpoint/i}}
haml :predict
end
get '/jme_help/?' do
File.read(File.join('views','jme_help.html'))
end
# get individual compound details
get '/prediction/:neighbor/details/?' do
@compound = OpenTox::Compound.new params[:neighbor]
@smiles = @compound.smiles
task = OpenTox::Task.run("Get names for '#{@smiles}'.") do
names = @compound.names
end
task.wait
case task[RDF::OT.hasStatus]
when "Error"
@names = "No names for this compound available."
when "Completed"
@names = @compound.names
else
@names = "No names for this compound available."
end
@inchi = @compound.inchi.gsub("InChI=", "")
haml :details, :layout => false
end
# fingerprints for compound in predictions
get '/prediction/:model_uri/:type/:compound_uri/fingerprints/?' do
@type = params[:type]
model = OpenTox::Model::Lazar.find params[:model_uri]
feature_dataset = OpenTox::Dataset.find model[RDF::OT.featureDataset]
@compound = OpenTox::Compound.new params[:compound_uri]
@significant_fragments = []
if @type =~ /classification/i
# collect all feature values with fingerprint
fingerprints = OpenTox::Algorithm::Descriptor.send("smarts_match", [@compound], feature_dataset.features.collect{ |f| f[RDF::DC.title]})[@compound.uri]
#$logger.debug "fingerprints:\t#{fingerprints}\n"
# collect fingerprints with value 1
@fingerprint_values = fingerprints.collect{|smarts, value| [smarts, value] if value > 0}
# collect all features from feature_dataset
@features = feature_dataset.features.collect{|f| f }
# search for each fingerprint in all features and collect feature values( effect, smarts, pValue )
@fingerprint_values.each{ |fi, v| @features.each{ |f| @significant_fragments << [f[RDF::OT.effect].to_i, f[RDF::OT.smarts], f[RDF::OT.pValue]] if fi == f[RDF::OT.smarts] } }
# pass value_map, important to interprete effect value
prediction_feature_uri = ""
model.parameters.each {|p|
if p[RDF::DC.title].to_s == "prediction_feature_uri"
prediction_feature_uri = p[RDF::OT.paramValue].object
end
}
prediction_feature = OpenTox::Feature.find prediction_feature_uri
@value_map = prediction_feature.value_map
else #regression
feature_calc_algo = ""
model.parameters.each {|p|
if p[RDF::DC.title].to_s == "feature_calculation_algorithm"
feature_calc_algo = p[RDF::OT.paramValue].object
end
}
@desc = []
fingerprints = OpenTox::Algorithm::Descriptor.send( feature_calc_algo, [ @compound ], feature_dataset.features.collect{ |f| f[RDF::DC.title] } )
fingerprints.each{|x, h| h.each{|descriptor, value| @desc << [descriptor, [value]]}}
pc_descriptor_titles_descriptions = {}
feature_dataset.features.collect{ |f|
pc_descriptor_titles_descriptions[f[RDF::DC.title]]= f[RDF::DC.description]
}
@desc.each{|d, v| @significant_fragments << [pc_descriptor_titles_descriptions[d], v] }
end
haml :significant_fragments, :layout => false
end
get '/prediction/:model_uri/:type/:neighbor/significant_fragments/?' do
@type = params[:type]
@compound = OpenTox::Compound.new params[:neighbor]
model = OpenTox::Model::Lazar.find params[:model_uri]
#$logger.debug "model for significant fragments:\t#{model.uri}"
feature_dataset = OpenTox::Dataset.find model[RDF::OT.featureDataset]
$logger.debug "feature_dataset_uri:\t#{feature_dataset.uri}\n"
# load all compounds
feature_dataset.compounds
# load all features
@features = feature_dataset.features.collect{|f| f}
# find all features and values for a neighbor compound
@significant_fragments = []
# check type first
if @type =~ /classification/i
# get compound index in feature dataset
c_idx = feature_dataset.compound_indices @compound.uri
# collect feature uris with value
@feat = @features.collect{|f| [feature_dataset.data_entry_value(c_idx[0], f.uri), f.uri]}
#$logger.debug "@feat:\t#{@feat}\n"
# pass feature uris if value > 0
@feat.each do |f|
# search relevant features
if f[0] > 0
f = OpenTox::Feature.find f[1]
# pass relevant features with [ effect, smarts, pValue ]
@significant_fragments << [f[RDF::OT.effect].to_i, f[RDF::OT.smarts], f[RDF::OT.pValue].to_f.round(3)]
end
end
# pass value_map, important to interprete effect value
prediction_feature_uri = ""
model.parameters.each {|p|
if p[RDF::DC.title].to_s == "prediction_feature_uri"
prediction_feature_uri = p[RDF::OT.paramValue].object
end
}
prediction_feature = OpenTox::Feature.find prediction_feature_uri
@value_map = prediction_feature.value_map
else # regression
# find a value in feature dataset by compound and feature
@values = @features.collect{|f| feature_dataset.values(@compound, f)}
#$logger.debug "values in fd:\t#{@values}"
@features.each_with_index{|f, i| @significant_fragments << [f.description, @values[i]]}
end
#$logger.debug "significant fragments:\t#{@significant_fragments}\n"
haml :significant_fragments, :layout => false
end
post '/predict/?' do
# validate identifier input
task = OpenTox::Task.run("Validate SMILES string.") do
# transfered input
@identifier = params[:identifier]
# get compound from SMILES
@compound = OpenTox::Compound.from_smiles @identifier.to_s
# validate SMILES by converting to INCHI
inchi = @compound.inchi
end
# necessary to wait for task
task.wait
# case task fails return message smiles invalid
# case task completed go ahead
case task[RDF::OT.hasStatus]
when "Error"
@error_report = "Attention, '#{params[:identifier]}' is not a valid SMILES string."
haml :error
when "Completed"
@identifier = params[:identifier]
@compound = OpenTox::Compound.from_smiles @identifier.to_s
# init arrays
@prediction_models = []
@predictions = []
@model_type = []
# get selected models
#TODO compare if model is selected by uri not title
params[:selection].each do |model|
# selected model = model[0]
# compare selected with all models
@@models.each do |m|
@prediction_models << m if m.title =~ /#{model[0]}/i
end
end
# predict with selected models
# one prediction in 'pa' array = OpenTox::Dataset
# all collected predictions in '@predictions' array
@prediction_models.each do |m|
# define type (classification|regression)
m.type.join =~ /classification/i ? (@model_type << "classification") : (@model_type << "regression")
# predict against compound
@prediction_uri = m.run :compound_uri => "#{@compound.uri}"
$logger.debug "prediction dataset:\t#{@prediction_uri}\n"
prediction = OpenTox::Dataset.new @prediction_uri
pa = []
pa << prediction
@predictions << pa
end
haml :prediction
end
end
get '/predict/stylesheets/:name.css' do
content_type 'text/css', :charset => 'utf-8'
sass(:"stylesheets/#{params[:name]}", Compass.sass_engine_options )
end
|
