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
237
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
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
346
347
348
349
350
351
352
353
354
355
356
357
358
359
360
361
362
363
364
365
366
367
368
369
370
371
372
373
374
375
376
377
378
379
380
381
382
383
384
385
386
387
388
389
390
391
392
393
394
395
396
397
398
399
400
401
402
403
404
405
406
407
408
409
410
411
412
413
414
415
416
417
418
419
420
421
422
423
424
425
426
427
428
429
430
431
432
433
434
435
436
437
438
439
440
441
442
443
444
445
446
447
448
449
450
451
452
453
454
455
456
457
458
459
460
461
462
463
464
465
466
467
468
469
470
471
472
473
474
475
476
477
478
479
480
481
482
483
484
485
486
487
488
489
490
491
492
493
494
495
496
497
498
499
500
501
502
503
504
505
506
507
508
509
510
|
module OpenTox
module Model
class Lazar
include OpenTox
include Mongoid::Document
include Mongoid::Timestamps
store_in collection: "models"
attr_writer :independent_variables # store in GridFS to avoid Mongo database size limit problems
field :name, type: String
field :creator, type: String, default: __FILE__
field :algorithms, type: Hash, default:{}
field :training_dataset_id, type: BSON::ObjectId
field :substance_ids, type: Array, default:[]
field :prediction_feature_id, type: BSON::ObjectId
field :dependent_variables, type: Array, default:[]
field :descriptor_ids, type:Array, default:[]
field :independent_variables_id, type: BSON::ObjectId
field :fingerprints, type: Array, default:[]
field :descriptor_weights, type: Array, default:[]
field :descriptor_means, type: Array, default:[]
field :descriptor_sds, type: Array, default:[]
field :scaled_variables, type: Array, default:[]
field :version, type: Hash, default:{}
# Create a lazar model
# @param [OpenTox::Dataset] training_dataset
# @param [OpenTox::Feature, nil] prediction_feature
# By default the first feature of the training dataset will be predicted, specify a prediction_feature if you want to predict another feature
# @param [Hash, nil] algorithms
# Default algorithms will be used, if no algorithms parameter is provided. The algorithms hash has the following keys: :descriptors (specifies the descriptors to be used for similarity calculations and local QSAR models), :similarity (similarity algorithm and threshold), :feature_selection (feature selection algorithm), :prediction (local QSAR algorithm). Default parameters are used for unspecified keys.
#
# @return [OpenTox::Model::Lazar]
def self.create prediction_feature:nil, training_dataset:, algorithms:{}
bad_request_error "Please provide a prediction_feature and/or a training_dataset." unless prediction_feature or training_dataset
prediction_feature = training_dataset.features.first unless prediction_feature
# TODO: prediction_feature without training_dataset: use all available data
# guess model type
prediction_feature.numeric? ? model = LazarRegression.new : model = LazarClassification.new
model.prediction_feature_id = prediction_feature.id
model.training_dataset_id = training_dataset.id
model.name = "#{prediction_feature.name} (#{training_dataset.name})"
# TODO: check if this works for gem version, add gem versioning?
dir = File.dirname(__FILE__)
commit = `cd #{dir}; git rev-parse HEAD`.chomp
branch = `cd #{dir}; git rev-parse --abbrev-ref HEAD`.chomp
url = `cd #{dir}; git config --get remote.origin.url`.chomp
if branch
model.version = {:url => url, :branch => branch, :commit => commit}
else
model.version = {:warning => "git is not installed"}
end
# set defaults#
substance_classes = training_dataset.substances.collect{|s| s.class.to_s}.uniq
bad_request_error "Cannot create models for mixed substance classes '#{substance_classes.join ', '}'." unless substance_classes.size == 1
if substance_classes.first == "OpenTox::Compound"
model.algorithms = {
:descriptors => {
:method => "fingerprint",
:type => "MP2D",
},
:feature_selection => nil
}
if model.class == LazarClassification
model.algorithms[:prediction] = {
:method => "Algorithm::Classification.weighted_majority_vote",
}
model.algorithms[:similarity] = {
:method => "Algorithm::Similarity.tanimoto",
:min => 0.1,
}
elsif model.class == LazarRegression
model.algorithms[:prediction] = {
:method => "Algorithm::Caret.rf",
}
model.algorithms[:similarity] = {
:method => "Algorithm::Similarity.tanimoto",
:min => 0.5,
}
end
elsif substance_classes.first == "OpenTox::Nanoparticle"
model.algorithms = {
:descriptors => {
:method => "properties",
:categories => ["P-CHEM"],
},
:similarity => {
:method => "Algorithm::Similarity.weighted_cosine",
:min => 0.5,
},
:prediction => {
:method => "Algorithm::Caret.rf",
},
:feature_selection => {
:method => "Algorithm::FeatureSelection.correlation_filter",
},
}
else
bad_request_error "Cannot create models for #{substance_classes.first}."
end
# overwrite defaults with explicit parameters
algorithms.each do |type,parameters|
if parameters and parameters.is_a? Hash
parameters.each do |p,v|
model.algorithms[type] ||= {}
model.algorithms[type][p] = v
model.algorithms[:descriptors].delete :categories if type == :descriptors and p == :type
end
else
model.algorithms[type] = parameters
end
end if algorithms
# parse dependent_variables from training dataset
training_dataset.substances.each do |substance|
values = training_dataset.values(substance,model.prediction_feature_id)
values.each do |v|
model.substance_ids << substance.id.to_s
model.dependent_variables << v
end if values
end
descriptor_method = model.algorithms[:descriptors][:method]
model.independent_variables = []
case descriptor_method
# parse fingerprints
when "fingerprint"
type = model.algorithms[:descriptors][:type]
model.substances.each_with_index do |s,i|
model.fingerprints[i] ||= []
model.fingerprints[i] += s.fingerprint(type)
model.fingerprints[i].uniq!
end
model.descriptor_ids = model.fingerprints.flatten.uniq
model.descriptor_ids.each do |d|
model.independent_variables << model.substance_ids.collect_with_index{|s,i| model.fingerprints[i].include? d} if model.algorithms[:prediction][:method].match /Caret/
end
# calculate physchem properties
when "calculate_properties"
features = model.algorithms[:descriptors][:features]
model.descriptor_ids = features.collect{|f| f.id.to_s}
model.algorithms[:descriptors].delete(:features)
model.algorithms[:descriptors].delete(:type)
model.substances.each_with_index do |s,i|
props = s.calculate_properties(features)
props.each_with_index do |v,j|
model.independent_variables[j] ||= []
model.independent_variables[j][i] = v
end if props and !props.empty?
end
# parse independent_variables
when "properties"
categories = model.algorithms[:descriptors][:categories]
feature_ids = []
categories.each do |category|
Feature.where(category:category).each{|f| feature_ids << f.id.to_s}
end
properties = model.substances.collect { |s| s.properties }
property_ids = properties.collect{|p| p.keys}.flatten.uniq
model.descriptor_ids = feature_ids & property_ids
model.independent_variables = model.descriptor_ids.collect{|i| properties.collect{|p| p[i] ? p[i].median : nil}}
else
bad_request_error "Descriptor method '#{descriptor_method}' not implemented."
end
if model.algorithms[:feature_selection] and model.algorithms[:feature_selection][:method]
model = Algorithm.run model.algorithms[:feature_selection][:method], model
end
# scale independent_variables
unless model.fingerprints?
model.independent_variables.each_with_index do |var,i|
model.descriptor_means[i] = var.mean
model.descriptor_sds[i] = var.standard_deviation
model.scaled_variables << var.collect{|v| v ? (v-model.descriptor_means[i])/model.descriptor_sds[i] : nil}
end
end
model.save
model
end
# Predict a substance (compound or nanoparticle)
# @param [OpenTox::Substance]
# @return [Hash]
def predict_substance substance, threshold = self.algorithms[:similarity][:min]
@independent_variables = Marshal.load $gridfs.find_one(_id: self.independent_variables_id).data
case algorithms[:similarity][:method]
when /tanimoto/ # binary features
similarity_descriptors = substance.fingerprint algorithms[:descriptors][:type]
# TODO this excludes descriptors only present in the query substance
# use for applicability domain?
query_descriptors = descriptor_ids.collect{|id| similarity_descriptors.include? id}
when /euclid|cosine/ # quantitative features
if algorithms[:descriptors][:method] == "calculate_properties" # calculate descriptors
features = descriptor_ids.collect{|id| Feature.find(id)}
query_descriptors = substance.calculate_properties(features)
similarity_descriptors = query_descriptors.collect_with_index{|v,i| (v-descriptor_means[i])/descriptor_sds[i]}
else
similarity_descriptors = []
query_descriptors = []
descriptor_ids.each_with_index do |id,i|
prop = substance.properties[id]
prop = prop.median if prop.is_a? Array # measured
if prop
similarity_descriptors[i] = (prop-descriptor_means[i])/descriptor_sds[i]
query_descriptors[i] = prop
end
end
end
else
bad_request_error "Unknown descriptor type '#{descriptors}' for similarity method '#{similarity[:method]}'."
end
prediction = {:warnings => [], :measurements => []}
prediction[:warnings] << "Similarity threshold #{threshold} < #{algorithms[:similarity][:min]}, prediction may be out of applicability domain." if threshold < algorithms[:similarity][:min]
neighbor_ids = []
neighbor_similarities = []
neighbor_dependent_variables = []
neighbor_independent_variables = []
# find neighbors
substance_ids.each_with_index do |s,i|
# handle query substance
if substance.id.to_s == s
prediction[:measurements] << dependent_variables[i]
prediction[:info] = "Substance '#{substance.name}, id:#{substance.id}' has been excluded from neighbors, because it is identical with the query substance."
else
if fingerprints?
neighbor_descriptors = fingerprints[i]
else
next if substance.is_a? Nanoparticle and substance.core != Nanoparticle.find(s).core # necessary for nanoparticle properties predictions
neighbor_descriptors = scaled_variables.collect{|v| v[i]}
end
sim = Algorithm.run algorithms[:similarity][:method], [similarity_descriptors, neighbor_descriptors, descriptor_weights]
if sim >= threshold
neighbor_ids << s
neighbor_similarities << sim
neighbor_dependent_variables << dependent_variables[i]
independent_variables.each_with_index do |c,j|
neighbor_independent_variables[j] ||= []
neighbor_independent_variables[j] << @independent_variables[j][i]
end
end
end
end
measurements = nil
if neighbor_similarities.empty?
prediction[:value] = nil
prediction[:warnings] << "Could not find similar substances with experimental data in the training dataset."
elsif neighbor_similarities.size == 1
prediction[:value] = nil
prediction[:warnings] << "Cannot create prediction: Only one similar compound in the training set."
prediction[:neighbors] = [{:id => neighbor_ids.first, :similarity => neighbor_similarities.first}]
else
query_descriptors.collect!{|d| d ? 1 : 0} if algorithms[:feature_selection] and algorithms[:descriptors][:method] == "fingerprint"
# call prediction algorithm
result = Algorithm.run algorithms[:prediction][:method], dependent_variables:neighbor_dependent_variables,independent_variables:neighbor_independent_variables ,weights:neighbor_similarities, query_variables:query_descriptors
prediction.merge! result
prediction[:neighbors] = neighbor_ids.collect_with_index{|id,i| {:id => id, :measurement => neighbor_dependent_variables[i], :similarity => neighbor_similarities[i]}}
#if neighbor_similarities.max < algorithms[:similarity][:warn_min]
#prediction[:warnings] << "Closest neighbor has similarity < #{algorithms[:similarity][:warn_min]}. Prediction may be out of applicability domain."
#end
end
if prediction[:warnings].empty? or threshold < algorithms[:similarity][:min] or threshold <= 0.2
prediction
else # try again with a lower threshold
predict_substance substance, 0.2
end
end
# Predict a substance (compound or nanoparticle), an array of substances or a dataset
# @param [OpenTox::Compound, OpenTox::Nanoparticle, Array<OpenTox::Substance>, OpenTox::Dataset]
# @return [Hash, Array<Hash>, OpenTox::Dataset]
def predict object
training_dataset = Dataset.find training_dataset_id
# parse data
substances = []
if object.is_a? Substance
substances = [object]
elsif object.is_a? Array
substances = object
elsif object.is_a? Dataset
substances = object.substances
else
bad_request_error "Please provide a OpenTox::Compound an Array of OpenTox::Substances or an OpenTox::Dataset as parameter."
end
# make predictions
predictions = {}
substances.each do |c|
predictions[c.id.to_s] = predict_substance c
predictions[c.id.to_s][:prediction_feature_id] = prediction_feature_id
end
# serialize result
if object.is_a? Substance
prediction = predictions[substances.first.id.to_s]
prediction[:neighbors].sort!{|a,b| b[1] <=> a[1]} if prediction[:neighbors]# sort according to similarity
return prediction
elsif object.is_a? Array
return predictions
elsif object.is_a? Dataset
# prepare prediction dataset
measurement_feature = Feature.find prediction_feature_id
prediction_feature = NumericFeature.find_or_create_by( "name" => measurement_feature.name + " (Prediction)" )
prediction_dataset = LazarPrediction.create(
:name => "Lazar prediction for #{prediction_feature.name}",
:creator => __FILE__,
:prediction_feature_id => prediction_feature.id,
:predictions => predictions
)
return prediction_dataset
end
end
# Save the model
# Stores independent_variables in GridFS to avoid Mongo database size limit problems
def save
file = Mongo::Grid::File.new(Marshal.dump(@independent_variables), :filename => "#{id}.independent_variables")
self.independent_variables_id = $gridfs.insert_one(file)
super
end
# Get independent variables
# @return [Array<Array>]
def independent_variables
@independent_variables ||= Marshal.load $gridfs.find_one(_id: self.independent_variables_id).data
@independent_variables
end
# Get training dataset
# @return [OpenTox::Dataset]
def training_dataset
Dataset.find(training_dataset_id)
end
# Get prediction feature
# @return [OpenTox::Feature]
def prediction_feature
Feature.find(prediction_feature_id)
end
# Get training descriptors
# @return [Array<OpenTox::Feature>]
def descriptors
descriptor_ids.collect{|id| Feature.find(id)}
end
# Get training substances
# @return [Array<OpenTox::Substance>]
def substances
substance_ids.collect{|id| Substance.find(id)}
end
# Are fingerprints used as descriptors
# @return [TrueClass, FalseClass]
def fingerprints?
algorithms[:descriptors][:method] == "fingerprint" ? true : false
end
end
# Classification model
class LazarClassification < Lazar
end
# Regression model
class LazarRegression < Lazar
end
# Convenience class for generating and validating lazar models in a single step and predicting substances (compounds and nanoparticles), arrays of substances and datasets
class Validation
include OpenTox
include Mongoid::Document
include Mongoid::Timestamps
field :endpoint, type: String
field :qmrf, type: Hash
field :species, type: String
field :source, type: String
field :unit, type: String
field :model_id, type: BSON::ObjectId
field :repeated_crossvalidation_id, type: BSON::ObjectId
# Predict a substance (compound or nanoparticle), an array of substances or a dataset
# @param [OpenTox::Compound, OpenTox::Nanoparticle, Array<OpenTox::Substance>, OpenTox::Dataset]
# @return [Hash, Array<Hash>, OpenTox::Dataset]
def predict object
model.predict object
end
# Get training dataset
# @return [OpenTox::Dataset]
def training_dataset
model.training_dataset
end
# Get lazar model
# @return [OpenTox::Model::Lazar]
def model
Lazar.find model_id
end
# Get algorithms
# @return [Hash]
def algorithms
model.algorithms
end
# Get prediction feature
# @return [OpenTox::Feature]
def prediction_feature
model.prediction_feature
end
# Get repeated crossvalidations
# @return [OpenTox::Validation::RepeatedCrossValidation]
def repeated_crossvalidation
OpenTox::Validation::RepeatedCrossValidation.find repeated_crossvalidation_id # full class name required
end
# Get crossvalidations
# @return [Array<OpenTox::CrossValidation]
def crossvalidations
repeated_crossvalidation.crossvalidations
end
# Is it a regression model
# @return [TrueClass, FalseClass]
def regression?
model.is_a? LazarRegression
end
# Is it a classification model
# @return [TrueClass, FalseClass]
def classification?
model.is_a? LazarClassification
end
# Create and validate a lazar model from a csv file with training data and a json file with metadata
# @param [File] CSV file with two columns. The first line should contain either SMILES or InChI (first column) and the endpoint (second column). The first column should contain either the SMILES or InChI of the training compounds, the second column the training compounds toxic activities (qualitative or quantitative). Use -log10 transformed values for regression datasets. Add metadata to a JSON file with the same basename containing the fields "species", "endpoint", "source" and "unit" (regression only). You can find example training data at https://github.com/opentox/lazar-public-data.
# @return [OpenTox::Model::Validation] lazar model with three independent 10-fold crossvalidations
def self.from_csv_file file
metadata_file = file.sub(/csv$/,"json")
bad_request_error "No metadata file #{metadata_file}" unless File.exist? metadata_file
model_validation = self.new JSON.parse(File.read(metadata_file))
training_dataset = Dataset.from_csv_file file
model = Lazar.create training_dataset: training_dataset
model_validation[:model_id] = model.id
model_validation[:repeated_crossvalidation_id] = OpenTox::Validation::RepeatedCrossValidation.create(model).id # full class name required
model_validation.save
model_validation
end
# Create and validate a nano-lazar model, import data from eNanoMapper if necessary
# nano-lazar methods are described in detail in https://github.com/enanomapper/nano-lazar-paper/blob/master/nano-lazar.pdf
# @param [OpenTox::Dataset, nil] training_dataset
# @param [OpenTox::Feature, nil] prediction_feature
# @param [Hash, nil] algorithms
# @return [OpenTox::Model::Validation] lazar model with five independent 10-fold crossvalidations
def self.from_enanomapper training_dataset: nil, prediction_feature:nil, algorithms: nil
# find/import training_dataset
training_dataset ||= Dataset.where(:name => "Protein Corona Fingerprinting Predicts the Cellular Interaction of Gold and Silver Nanoparticles").first
unless training_dataset # try to import
Import::Enanomapper.import
training_dataset = Dataset.where(name: "Protein Corona Fingerprinting Predicts the Cellular Interaction of Gold and Silver Nanoparticles").first
bad_request_error "Cannot import 'Protein Corona Fingerprinting Predicts the Cellular Interaction of Gold and Silver Nanoparticles' dataset" unless training_dataset
end
prediction_feature ||= Feature.where(name: "log2(Net cell association)", category: "TOX").first
model_validation = self.new(
:endpoint => prediction_feature.name,
:source => prediction_feature.source,
:species => "A549 human lung epithelial carcinoma cells",
:unit => prediction_feature.unit
)
model = LazarRegression.create prediction_feature: prediction_feature, training_dataset: training_dataset, algorithms: algorithms
model_validation[:model_id] = model.id
repeated_cv = OpenTox::Validation::RepeatedCrossValidation.create model, 10, 5
model_validation[:repeated_crossvalidation_id] = repeated_cv.id
model_validation.save
model_validation
end
end
end
end
|