5 * Created by westcott on 11/3/09.
6 * Copyright 2009 Schloss Lab. All rights reserved.
13 /**************************************************************************************************/
14 Bayesian::Bayesian(string tfile, string tempFile, string method, int ksize, int cutoff, int i) :
15 Classify(tfile, tempFile, method, ksize, 0.0, 0.0, 0.0, 0.0), kmerSize(ksize), confidenceThreshold(cutoff), iters(i) {
18 numKmers = database->getMaxKmer() + 1;
20 //initialze probabilities
21 wordGenusProb.resize(numKmers);
23 genusNodes = phyloTree->getGenusNodes();
25 for (int j = 0; j < wordGenusProb.size(); j++) { wordGenusProb[j].resize(genusNodes.size()); }
27 //reset counts because we are on a new word
28 for (int j = 0; j < genusNodes.size(); j++) {
29 TaxNode temp = phyloTree->get(genusNodes[j]);
30 genusTotals.push_back(temp.accessions.size());
34 /************calculate the probablity that each word will be in a specific taxonomy*************/
36 string probFileName = tfile.substr(0,tfile.find_last_of(".")+1) + tempFile.substr(0,tempFile.find_last_of(".")+1) + char('0'+ kmerSize) + "mer.prob";
37 ifstream probFileTest(probFileName.c_str());
40 string probFileName2 = tfile.substr(0,tfile.find_last_of(".")+1) + tempFile.substr(0,tempFile.find_last_of(".")+1) + char('0'+ kmerSize) + "mer.numNonZero";
41 ifstream probFileTest2(probFileName2.c_str());
43 int start = time(NULL);
45 if(probFileTest && probFileTest2){
46 m->mothurOut("Reading template probabilities... "); cout.flush();
47 readProbFile(probFileTest, probFileTest2);
49 m->mothurOut("Calculating template probabilities... "); cout.flush();
52 openOutputFile(probFileName, out);
55 openOutputFile(probFileName2, out2);
58 for (int i = 0; i < numKmers; i++) {
59 if (m->control_pressed) { break; }
63 vector<int> seqsWithWordi = database->getSequencesWithKmer(i);
66 for (int k = 0; k < genusNodes.size(); k++) { count[genusNodes[k]] = 0; }
68 //for each sequence with that word
69 for (int j = 0; j < seqsWithWordi.size(); j++) {
70 int temp = phyloTree->getIndex(names[seqsWithWordi[j]]);
71 count[temp]++; //increment count of seq in this genus who have this word
74 //probabilityInTemplate = (# of seqs with that word in template + 0.05) / (total number of seqs in template + 1);
75 float probabilityInTemplate = (seqsWithWordi.size() + 0.50) / (float) (names.size() + 1);
78 for (int k = 0; k < genusNodes.size(); k++) {
79 //probabilityInThisTaxonomy = (# of seqs with that word in this taxonomy + probabilityInTemplate) / (total number of seqs in this taxonomy + 1);
80 wordGenusProb[i][k] = log((count[genusNodes[k]] + probabilityInTemplate) / (float) (genusTotals[k] + 1));
81 if (count[genusNodes[k]] != 0) { out << k << '\t' << wordGenusProb[i][k] << '\t'; numNotZero++; }
84 out2 << probabilityInTemplate << '\t' << numNotZero << endl;
92 m->mothurOut("DONE."); m->mothurOutEndLine();
93 m->mothurOut("It took " + toString(time(NULL) - start) + " seconds get probabilities. "); m->mothurOutEndLine();
96 m->errorOut(e, "Bayesian", "Bayesian");
100 /**************************************************************************************************/
101 string Bayesian::getTaxonomy(Sequence* seq) {
106 //get words contained in query
107 //getKmerString returns a string where the index in the string is hte kmer number
108 //and the character at that index can be converted to be the number of times that kmer was seen
110 string queryKmerString = kmer.getKmerString(seq->getUnaligned());
112 vector<int> queryKmers;
113 for (int i = 0; i < queryKmerString.length(); i++) {
114 if (queryKmerString[i] != '!') { //this kmer is in the query
115 queryKmers.push_back(i);
120 if (queryKmers.size() == 0) { m->mothurOut(seq->getName() + "is bad."); m->mothurOutEndLine(); return "bad seq"; }
122 int index = getMostProbableTaxonomy(queryKmers);
125 if (m->control_pressed) { return tax; }
127 //bootstrap - to set confidenceScore
128 int numToSelect = queryKmers.size() / 8;
129 tax = bootstrapResults(queryKmers, index, numToSelect);
133 catch(exception& e) {
134 m->errorOut(e, "Bayesian", "getTaxonomy");
138 /**************************************************************************************************/
139 string Bayesian::bootstrapResults(vector<int> kmers, int tax, int numToSelect) {
142 //taxConfidenceScore.clear(); //clear out previous seqs scores
144 vector< map<string, int> > confidenceScores; //you need the added vector level of confusion to account for the level that name is seen since they can be the same
145 //map of classification to confidence for all areas seen
146 //ie. Bacteria;Alphaproteobacteria;Rhizobiales;Azorhizobium_et_rel.;Methylobacterium_et_rel.;Bosea;
147 //ie. Bacteria -> 100, Alphaproteobacteria -> 100, Rhizobiales -> 87, Azorhizobium_et_rel. -> 78, Methylobacterium_et_rel. -> 70, Bosea -> 50
148 confidenceScores.resize(100); //if you have more than 100 levels of classification...
150 map<string, int>::iterator itBoot;
151 map<string, int>::iterator itBoot2;
152 map<int, int>::iterator itConvert;
154 for (int i = 0; i < iters; i++) {
155 if (m->control_pressed) { return "control"; }
159 for (int j = 0; j < numToSelect; j++) {
160 int index = int(rand() % kmers.size());
163 temp.push_back(kmers[index]);
167 int newTax = getMostProbableTaxonomy(temp);
168 TaxNode taxonomy = phyloTree->get(newTax);
170 //add to confidence results
171 while (taxonomy.level != 0) { //while you are not at the root
173 itBoot2 = confidenceScores[taxonomy.level].find(taxonomy.name); //is this a classification we already have a count on
175 if (itBoot2 == confidenceScores[taxonomy.level].end()) { //not already in confidence scores
176 confidenceScores[taxonomy.level][taxonomy.name] = 1;
178 confidenceScores[taxonomy.level][taxonomy.name]++;
181 taxonomy = phyloTree->get(taxonomy.parent);
186 string confidenceTax = "";
188 TaxNode seqTax = phyloTree->get(tax);
190 while (seqTax.level != 0) { //while you are not at the root
192 itBoot2 = confidenceScores[seqTax.level].find(seqTax.name); //is this a classification we already have a count on
195 if (itBoot2 != confidenceScores[seqTax.level].end()) { //not already in confidence scores
196 confidence = confidenceScores[seqTax.level][seqTax.name];
199 if (confidence >= confidenceThreshold) {
200 confidenceTax = seqTax.name + "(" + toString(((confidence/(float)iters) * 100)) + ");" + confidenceTax;
201 simpleTax = seqTax.name + ";" + simpleTax;
204 seqTax = phyloTree->get(seqTax.parent);
207 if (confidenceTax == "") { confidenceTax = "unclassified;"; simpleTax = "unclassified;"; }
208 return confidenceTax;
211 catch(exception& e) {
212 m->errorOut(e, "Bayesian", "bootstrapResults");
216 /**************************************************************************************************/
217 int Bayesian::getMostProbableTaxonomy(vector<int> queryKmer) {
219 int indexofGenus = 0;
221 double maxProbability = -1000000.0;
222 //find taxonomy with highest probability that this sequence is from it
223 for (int k = 0; k < genusNodes.size(); k++) {
224 //for each taxonomy calc its probability
226 for (int i = 0; i < queryKmer.size(); i++) {
227 prob += wordGenusProb[queryKmer[i]][k];
230 //is this the taxonomy with the greatest probability?
231 if (prob > maxProbability) {
232 indexofGenus = genusNodes[k];
233 maxProbability = prob;
239 catch(exception& e) {
240 m->errorOut(e, "Bayesian", "getMostProbableTaxonomy");
244 /*************************************************************************************************
245 map<string, int> Bayesian::parseTaxMap(string newTax) {
248 map<string, int> parsed;
250 newTax = newTax.substr(0, newTax.length()-1); //get rid of last ';'
254 while (newTax.find_first_of(';') != -1) {
255 individual = newTax.substr(0,newTax.find_first_of(';'));
256 newTax = newTax.substr(newTax.find_first_of(';')+1, newTax.length());
257 parsed[individual] = 1;
266 catch(exception& e) {
267 m->errorOut(e, "Bayesian", "parseTax");
271 /**************************************************************************************************/
272 void Bayesian::readProbFile(ifstream& in, ifstream& inNum) {
275 int kmer, name, count; count = 0;
276 vector<int> num; num.resize(numKmers);
278 vector<float> zeroCountProb; zeroCountProb.resize(numKmers);
281 inNum >> zeroCountProb[count] >> num[count];
290 //set them all to zero value
291 for (int i = 0; i < genusNodes.size(); i++) {
292 wordGenusProb[kmer][i] = log(zeroCountProb[kmer] / (float) (genusTotals[i]+1));
295 //get probs for nonzero values
296 for (int i = 0; i < num[kmer]; i++) {
298 wordGenusProb[kmer][name] = prob;
305 catch(exception& e) {
306 m->errorOut(e, "Bayesian", "readProbFile");
310 /**************************************************************************************************/