Package edu.msu.cme.rdp.classifier.train

Examples of edu.msu.cme.rdp.classifier.train.LineageSequenceParser

37
38
39
40
41
42
43
44
45
46
47
48
49
         // create a tree
        createTree(factory, new File(inFile));                           
        outWriter = new BufferedWriter(new FileWriter(outFile));
        outWriter.write("taxon file: " + taxFile + "\n" + "train sequence file: " + inFile +"\n");       
       
        LineageSequenceParser parser = new LineageSequenceParser(new File(testFile) );
        while ( parser.hasNext()){
            seqList.add(parser.next());
        }
       
        outWriter.write("query sequence file: " + testFile + "\n" + "classify moving window of size " + FindWindowFrame.window_size + ", step " + FindWindowFrame.step +"\n");
              
        Iterator windowIt = windowFrames.iterator();
View Full Code Here
70
71
72
73
74
75
76
77
78
79
80
81
82
83
    }
   
    /** reads from the stream, parses the sequence and creates the tree */
    private void createTree(TreeFactory factory, File input) throws IOException{
      long startTime = System.currentTimeMillis();
      LineageSequenceParser parser = new LineageSequenceParser(input);    
     
      while ( parser.hasNext() ){
        factory.addSequence( (LineageSequence)parser.next());
       
      }  
      //after all the training set is being parsed, calculate the prior probability for all the words.
      factory.calculateWordPrior();
      // create the word occurrence for all the nodes
View Full Code Here
77
78
79
80
81
82
83
84
85
86
87
        TreeFactory factory = new TreeFactory(new FileReader(taxFile));
        // create a tree
        createTree(factory, trainseqFile);

        BufferedWriter outWriter = new BufferedWriter(new FileWriter(outFile));
        LineageSequenceParser parser = new LineageSequenceParser(new File(testFile));
        LeaveOneOutTester tester = new LeaveOneOutTester(outWriter, numGoodBases);

        outWriter.write("taxon file: " + taxFile + "\n" + "train sequence file: " + trainseqFile + "\n");
        outWriter.write("word size: " + GoodWordIterator.getWordsize() + "\n");
        outWriter.write("minimum number of words for bootstrap: " + min_bootstrap_words + "\n");
View Full Code Here
101
102
103
104
105
106
107
108
109
110
111
112
113
114
    }

    /** reads from the stream, parses the sequences and creates the tree */
    private void createTree(TreeFactory factory, String input) throws IOException {
        LineageSequenceParser parser = new LineageSequenceParser(new File(input));

        while (parser.hasNext()) {
            factory.addSequence((LineageSequence) parser.next());
        }

        //after all the training set is being parsed, calculate the prior probability for all the words.
        factory.calculateWordPrior();
        // create the word occurrence for all the nodes, this is necessary if test on different level
View Full Code Here
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
   
    public void calSabSimilarity(String taxonFile, String trainSeqFile, String testSeqFile) throws IOException{       
        TreeFactory factory = new TreeFactory(new FileReader(taxonFile));
        factory.buildTree();
        // get the lineage of the trainSeqFile 
        LineageSequenceParser trainParser = new LineageSequenceParser(new File(trainSeqFile));
        HashMap<String, List<String>> lineageMap = new HashMap<String, List<String>>();
        while (trainParser.hasNext()) {
            LineageSequence seq = (LineageSequence) trainParser.next();
            lineageMap.put(seq.getSeqName(), seq.getAncestors());
           
         }
        trainParser.close();
        NuclSeqMatch sabCal = new NuclSeqMatch(trainSeqFile);
        LineageSequenceParser parser = new LineageSequenceParser(new File(testSeqFile));

        int count = 0;
        while (parser.hasNext()) {
            LineageSequence seq = (LineageSequence) parser.next();
            HashMap<String,HierarchyTree> queryAncestorNodes = getAncestorNodes(factory.getRoot(), seq.getSeqName(), seq.getAncestors());
           TreeSet<KmerMatchCore.BestMatch> matchResults = sabCal.findAllMatches(seq);
           
            short withinLowestRankSab = -1;
            short diffLowestRankSab = -1; 
            String bestDiffLowestRankMatch = null;
            for (KmerMatchCore.BestMatch match: matchResults){
                if ( match.getBestMatch().getSeqName().equals(seq.getSeqName())) continue;
                short sab = (short)(Math.round(100*match.getSab()));
                HashMap<String,HierarchyTree> matchAncestorNodes = getAncestorNodes(factory.getRoot(), match.getBestMatch().getSeqName(), lineageMap.get(match.getBestMatch().getSeqName()));
                boolean withinTaxon = false;
                for (int i = ranks.size() -1; i >=0; i--){                   
                    HierarchyTree queryTaxon = queryAncestorNodes.get( ranks.get(i));
                    HierarchyTree matchTaxon = matchAncestorNodes.get( ranks.get(i));
                    if ( queryTaxon != null && matchTaxon != null){
                        if ( queryTaxon.getName().equals(matchTaxon.getName())){
                            if ( !withinTaxon){  // if the query and match are not in the same child taxon, add sab to the current taxon
                                (sabCoutMap.get(ranks.get(i)))[sab]++;
                            }
                            withinTaxon = true;                           
                        }else {
                            withinTaxon = false;
                        }
                    }
                   
                } 
               
                // find within or different lowest level rank sab score, be either species or genus or any rank
                HierarchyTree speciesQueryTaxon = queryAncestorNodes.get( ranks.get(ranks.size()-1));   
                HierarchyTree speciesMatchTaxon = matchAncestorNodes.get( ranks.get(ranks.size()-1));
               
                if ( speciesQueryTaxon != null && speciesMatchTaxon != null && speciesQueryTaxon.getName().equals(speciesMatchTaxon.getName())){
                    withinLowestRankSab = sab >= withinLowestRankSab ? sab: withinLowestRankSab;
                }else {
                   
                    if ( sab >= diffLowestRankSab ){
                        bestDiffLowestRankMatch = match.getBestMatch().getSeqName();
                        diffLowestRankSab = sab;
                    }
                }
            }
            if ( withinLowestRankSab > 0){
                withinLowestRankSabSet.add(withinLowestRankSab);
            }
            if ( diffLowestRankSab > 0 ){
                diffLowestRankSabSet.add(diffLowestRankSab);
            }
            //System.out.println(seq.getSeqName() + "\t" + withinLowestRankSab + "\t" + diffLowestRankSab );
            count++;
            if ( count % 100 == 0){
                System.out.println(count);
            }
        }
        parser.close();
   
    }
View Full Code Here
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
   
    public void calPairwiseSimilaritye(String taxonFile, String trainSeqFile, String testSeqFile) throws IOException, OverlapCheckFailedException{       
        TreeFactory factory = new TreeFactory(new FileReader(taxonFile));
        factory.buildTree();
        // get the lineage of the trainSeqFile 
        LineageSequenceParser trainParser = new LineageSequenceParser(new File(trainSeqFile));
        ArrayList<LineageSequence> trainSeqList = new ArrayList<LineageSequence>();
        while (trainParser.hasNext()) {
            LineageSequence seq = (LineageSequence) trainParser.next();
            trainSeqList.add(seq);
         }
        trainParser.close();
        LineageSequenceParser parser = new LineageSequenceParser(new File(testSeqFile));

        while (parser.hasNext()) {
            LineageSequence seq = (LineageSequence) parser.next();
            HashMap<String,HierarchyTree> queryAncestorNodes = getAncestorNodes(factory.getRoot(), seq.getSeqName(), seq.getAncestors());
           
            for (LineageSequence trainSeq: trainSeqList){
                if ( trainSeq.getSeqName().equals(seq.getSeqName())) continue;
               
                HashMap<String,HierarchyTree> matchAncestorNodes = getAncestorNodes(factory.getRoot(), trainSeq.getSeqName(), trainSeq.getAncestors());
                boolean withinTaxon = false;
                String lowestCommonRank = null;
                for (int i = ranks.size() -1; i >=0; i--){                   
                    HierarchyTree queryTaxon = queryAncestorNodes.get( ranks.get(i));
                    HierarchyTree matchTaxon = matchAncestorNodes.get( ranks.get(i));
                    if ( queryTaxon != null && matchTaxon != null){
                        if ( queryTaxon.getName().equals(matchTaxon.getName())){
                            if ( !withinTaxon){  // if the query and match are not in the same child taxon, add sab to the current taxon
                                lowestCommonRank = ranks.get(i);
                                //(sabCoutMap.get(ranks.get(i)))[sab]++;
                            }
                            withinTaxon = true;                           
                        }else {
                            withinTaxon = false;
                        }
                    }
                } 
               
                if ( lowestCommonRank == null){  // not the rank we care
                    continue;
                }

                // we need to use overlap_trim mode and calculate distance as metric to count insertions, deletions and mismatches.
                PairwiseAlignment result = PairwiseAligner.align(seq.getSeqString().replaceAll("U", "T"), trainSeq.getSeqString().replaceAll("U", "T"), scoringMatrix, mode);
                short sab = (short) (100 - 100*dist.getDistance(result.getAlignedSeqj().getBytes(), result.getAlignedSeqi().getBytes(), 0));
                sabCoutMap.get(lowestCommonRank)[sab]++;
                               
            }          
        }
        parser.close();
   
    }
View Full Code Here
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
             }
         }

        int totalTest = 0;
        int totalSeq = 0;
        LineageSequenceParser parser = new LineageSequenceParser(source_file);
        while ( parser.hasNext()){
            totalSeq ++;
          LineageSequence pSeq = parser.next();
          if ( !selectedTestSeqIDs.contains(pSeq.getSeqName()) || pSeq.getSeqString().length() == 0){
              continue;
          }
          GoodWordIterator wordIterator = null ;
          if ( partialLength != null ){
                wordIterator = pSeq.getPartialSeqIteratorbyGoodBases(partialLength.intValue());  // test partial sequences with good words only

          }else {
                wordIterator = new GoodWordIterator(pSeq.getSeqString()); // full sequence
          }

          if (wordIterator == null || wordIterator.getNumofWords() == 0){
            //System.err.println(pSeq.getSeqName() + " unable to find good sequence");
            continue;
          }
       
          List result = dm.getBestClasspath( wordIterator, genusNodeMap, useSeed, min_bootstrap_words);

          //xxx
          ValidClassificationResultFacade resultFacade = new ValidClassificationResultFacade(pSeq, result);

          compareClassificationResult(factory, resultFacade, rankNodeMap, statusCountList);

          totalTest++;
        }
        parser.close();

        outWriter.write("taxon file\t" + tax_file.getName() + "\n" + "train sequence file\t" + source_file.getName() + "\n");
        outWriter.write("word size\t" + GoodWordIterator.getWordsize() + "\n");
        outWriter.write("minimum number of words for bootstrap\t" + min_bootstrap_words + "\n");
        if ( partialLength != null){
View Full Code Here
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
     * @throws IOException
     */
    private TreeFactory setup(File tax_file, File source_file, Set<String> selectedTestSeqIDs) throws IOException {

        TreeFactory factory = new TreeFactory(new FileReader(tax_file));
        LineageSequenceParser parser = new LineageSequenceParser(source_file);

        while ( parser.hasNext() ){
            LineageSequence pSeq = parser.next();
            if ( !selectedTestSeqIDs.contains(pSeq.getSeqName())){
              factory.addSequence( pSeq);
            }
        }
        parser.close();

        //after all the training set is being parsed, calculate the prior probability for all the words.
        factory.calculateWordPrior();
        return factory;
      }
View Full Code Here
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
        }
        // need to use ISO encoding for UNITE
        FileReader tax  = new FileReader(new File(taxFile));
        TreeFactory factory = new TreeFactory(tax);
              
        LineageSequenceParser parser = new LineageSequenceParser(new File(seqFile));
        LineageSequence seq;
        HashMap<String, String> seqMap = new HashMap<String, String>(); // seqID, desc
        while ( parser.hasNext()){
            seq = parser.next();
            factory.addSequence(seq, false); // donot check the kmers
           
            if ( seq.getSeqName().contains("|S00") ){ // rdpID
                String[] values = seq.getSeqName().split("\\|");
                seqMap.put(values[0], seq.getDesc());  
            }else if (seq.getSeqName().contains("|SH") ){  // if it's seq from UNITE, we need to do something with the seqID            
                String[] values = seq.getSeqName().split("\\|");
                seqMap.put(values[1], seq.getDesc());
            }else {
                seqMap.put(seq.getSeqName(), seq.getDesc());
            }
           
        }
        parser.close();
        HierarchyTreeExtend retVal = new HierarchyTreeExtend(factory.getRoot(), trainsetName);
        retVal.seqMap = seqMap;
        return retVal;
       
    }
View Full Code Here
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
   
    public PairwiseSeqDistance(String trainseqFile, String taxFile, AlignmentMode mode, boolean show_alignment ) throws IOException, OverlapCheckFailedException{
        this.mode = mode;
        this.show_alignment = show_alignment;
        factory = new TreeFactory(new FileReader(taxFile));
        LineageSequenceParser parser = new LineageSequenceParser(new File(trainseqFile));
       
        while (parser.hasNext()) {
            LineageSequence tmp = (LineageSequence) parser.next();
            seqList.add(tmp);
            factory.addSequence((LineageSequence) parser.next());
        }
        parser.close();
       
        calDist();
    }
View Full Code Here
TOP

Related Classes of edu.msu.cme.rdp.classifier.train.LineageSequenceParser

  • edu.msu.cme.rdp.classifier.train.validation.crossvalidate.CrossValidate
  • edu.msu.cme.rdp.classifier.train.validation.crossvalidate.RdmSelectTaxon
  • edu.msu.cme.rdp.classifier.train.validation.DecisionMakerTest
  • edu.msu.cme.rdp.classifier.train.validation.distance.CompareTrainingSets
  • edu.msu.cme.rdp.classifier.train.validation.distance.PairwiseSeqDistance
  • edu.msu.cme.rdp.classifier.train.validation.distance.TaxaSimilarityMain
  • edu.msu.cme.rdp.classifier.train.validation.HierarchyTreeTest
  • edu.msu.cme.rdp.classifier.train.validation.leaveoneout.LeaveOneOutTesterMain
  • edu.msu.cme.rdp.classifier.train.validation.movingwindow.MainMovingWindow
  • edu.msu.cme.rdp.classifier.train.validation.NBClassifierTest
    • Copyright © 2018 www.massapicom. All rights reserved.
    All source code are property of their respective owners. Java is a trademark of Sun Microsystems, Inc and owned by ORACLE Inc. Contact coftware#gmail.com.