X-Git-Url: https://git.donarmstrong.com/?a=blobdiff_plain;f=bayesian.cpp;h=b46f7703a971b8b5141d42faf14e26d45037cba0;hb=9946a1b4b50969d08ce059b248bdeecafbf989ac;hp=caae212f4855367653fef715e64a551eb4d095e9;hpb=63e089e0b3aad1741bab60119ed7ccc784dce347;p=mothur.git diff --git a/bayesian.cpp b/bayesian.cpp index caae212..b46f770 100644 --- a/bayesian.cpp +++ b/bayesian.cpp @@ -8,22 +8,524 @@ */ #include "bayesian.h" +#include "kmer.hpp" +#include "phylosummary.h" /**************************************************************************************************/ -Bayesian::Bayesian(string tfile, string tempFile, string method, int kmerSize, int gapOpen, int gapExtend, int match, int misMatch) : -Classify(tfile, tempFile, method, kmerSize, gapOpen, gapExtend, match, misMatch) {} +Bayesian::Bayesian(string tfile, string tempFile, string method, int ksize, int cutoff, int i) : +Classify(), kmerSize(ksize), confidenceThreshold(cutoff), iters(i) { + try { + + /************calculate the probablity that each word will be in a specific taxonomy*************/ + string tfileroot = tfile.substr(0,tfile.find_last_of(".")+1); + string tempfileroot = getRootName(getSimpleName(tempFile)); + string phyloTreeName = tfileroot + "tree.train"; + string probFileName = tfileroot + tempfileroot + char('0'+ kmerSize) + "mer.prob"; + string probFileName2 = tfileroot + tempfileroot + char('0'+ kmerSize) + "mer.numNonZero"; + + ofstream out; + ofstream out2; + + ifstream phyloTreeTest(phyloTreeName.c_str()); + ifstream probFileTest2(probFileName2.c_str()); + ifstream probFileTest(probFileName.c_str()); + + int start = time(NULL); + + if(probFileTest && probFileTest2 && phyloTreeTest){ + m->mothurOut("Reading template taxonomy... "); cout.flush(); + + phyloTree = new PhyloTree(phyloTreeTest, phyloTreeName); + + m->mothurOut("DONE."); m->mothurOutEndLine(); + + genusNodes = phyloTree->getGenusNodes(); + genusTotals = phyloTree->getGenusTotals(); + + m->mothurOut("Reading template probabilities... "); cout.flush(); + readProbFile(probFileTest, probFileTest2, probFileName, probFileName2); + + }else{ + + //create search database and names vector + generateDatabaseAndNames(tfile, tempFile, method, ksize, 0.0, 0.0, 0.0, 0.0); + + genusNodes = phyloTree->getGenusNodes(); + genusTotals = phyloTree->getGenusTotals(); + + m->mothurOut("Calculating template taxonomy tree... "); cout.flush(); + + phyloTree->printTreeNodes(phyloTreeName); + + m->mothurOut("DONE."); m->mothurOutEndLine(); + + m->mothurOut("Calculating template probabilities... "); cout.flush(); + + numKmers = database->getMaxKmer() + 1; + + //initialze probabilities + wordGenusProb.resize(numKmers); + + for (int j = 0; j < wordGenusProb.size(); j++) { wordGenusProb[j].resize(genusNodes.size()); } + + + #ifdef USE_MPI + int pid; + MPI_Comm_rank(MPI_COMM_WORLD, &pid); //find out who we are + + if (pid == 0) { + #endif + + ofstream out; + openOutputFile(probFileName, out); + + out << numKmers << endl; + + ofstream out2; + openOutputFile(probFileName2, out2); + + #ifdef USE_MPI + } + #endif + + + //for each word + for (int i = 0; i < numKmers; i++) { + if (m->control_pressed) { break; } + + #ifdef USE_MPI + MPI_Comm_rank(MPI_COMM_WORLD, &pid); //find out who we are + + if (pid == 0) { + #endif + + out << i << '\t'; + + #ifdef USE_MPI + } + #endif + + vector seqsWithWordi = database->getSequencesWithKmer(i); + + map count; + for (int k = 0; k < genusNodes.size(); k++) { count[genusNodes[k]] = 0; } + + //for each sequence with that word + for (int j = 0; j < seqsWithWordi.size(); j++) { + int temp = phyloTree->getIndex(names[seqsWithWordi[j]]); + count[temp]++; //increment count of seq in this genus who have this word + } + + //probabilityInTemplate = (# of seqs with that word in template + 0.05) / (total number of seqs in template + 1); + float probabilityInTemplate = (seqsWithWordi.size() + 0.50) / (float) (names.size() + 1); + + int numNotZero = 0; + for (int k = 0; k < genusNodes.size(); k++) { + //probabilityInThisTaxonomy = (# of seqs with that word in this taxonomy + probabilityInTemplate) / (total number of seqs in this taxonomy + 1); + wordGenusProb[i][k] = log((count[genusNodes[k]] + probabilityInTemplate) / (float) (genusTotals[k] + 1)); + if (count[genusNodes[k]] != 0) { + #ifdef USE_MPI + MPI_Comm_rank(MPI_COMM_WORLD, &pid); //find out who we are + if (pid == 0) { + #endif + + out << k << '\t' << wordGenusProb[i][k] << '\t'; + + #ifdef USE_MPI + } + #endif + + numNotZero++; + } + } + + #ifdef USE_MPI + MPI_Comm_rank(MPI_COMM_WORLD, &pid); //find out who we are + if (pid == 0) { + #endif + + out << endl; + out2 << probabilityInTemplate << '\t' << numNotZero << endl; + + #ifdef USE_MPI + } + #endif + } + + #ifdef USE_MPI + MPI_Comm_rank(MPI_COMM_WORLD, &pid); //find out who we are + if (pid == 0) { + #endif + + out.close(); + out2.close(); + + #ifdef USE_MPI + } + #endif + + //read in new phylotree with less info. - its faster + ifstream phyloTreeTest(phyloTreeName.c_str()); + delete phyloTree; + + phyloTree = new PhyloTree(phyloTreeTest, phyloTreeName); + } + + m->mothurOut("DONE."); m->mothurOutEndLine(); + m->mothurOut("It took " + toString(time(NULL) - start) + " seconds get probabilities. "); m->mothurOutEndLine(); + } + catch(exception& e) { + m->errorOut(e, "Bayesian", "Bayesian"); + exit(1); + } +} +/**************************************************************************************************/ +Bayesian::~Bayesian() { + try { + delete phyloTree; + if (database != NULL) { delete database; } + } + catch(exception& e) { + m->errorOut(e, "Bayesian", "~Bayesian"); + exit(1); + } +} + /**************************************************************************************************/ string Bayesian::getTaxonomy(Sequence* seq) { try { - string tax; + string tax = ""; + Kmer kmer(kmerSize); - + //get words contained in query + //getKmerString returns a string where the index in the string is hte kmer number + //and the character at that index can be converted to be the number of times that kmer was seen + + string queryKmerString = kmer.getKmerString(seq->getUnaligned()); + + vector queryKmers; + for (int i = 0; i < queryKmerString.length(); i++) { + if (queryKmerString[i] != '!') { //this kmer is in the query + queryKmers.push_back(i); + } + } + + if (queryKmers.size() == 0) { m->mothurOut(seq->getName() + "is bad."); m->mothurOutEndLine(); return "bad seq"; } + + int index = getMostProbableTaxonomy(queryKmers); + + if (m->control_pressed) { return tax; } +//cout << seq->getName() << '\t' << index << endl; + //bootstrap - to set confidenceScore + int numToSelect = queryKmers.size() / 8; + tax = bootstrapResults(queryKmers, index, numToSelect); + return tax; } catch(exception& e) { - errorOut(e, "Bayesian", "getTaxonomy"); + m->errorOut(e, "Bayesian", "getTaxonomy"); + exit(1); + } +} +/**************************************************************************************************/ +string Bayesian::bootstrapResults(vector kmers, int tax, int numToSelect) { + try { + + map confidenceScores; + + map::iterator itBoot; + map::iterator itBoot2; + map::iterator itConvert; + + for (int i = 0; i < iters; i++) { + if (m->control_pressed) { return "control"; } + + vector temp; + + for (int j = 0; j < numToSelect; j++) { + int index = int(rand() % kmers.size()); + + //add word to temp + temp.push_back(kmers[index]); + } + + //get taxonomy + int newTax = getMostProbableTaxonomy(temp); + TaxNode taxonomyTemp = phyloTree->get(newTax); + + //add to confidence results + while (taxonomyTemp.level != 0) { //while you are not at the root + + itBoot2 = confidenceScores.find(newTax); //is this a classification we already have a count on + + if (itBoot2 == confidenceScores.end()) { //not already in confidence scores + confidenceScores[newTax] = 1; + }else{ + confidenceScores[newTax]++; + } + + newTax = taxonomyTemp.parent; + taxonomyTemp = phyloTree->get(newTax); + } + + } + + string confidenceTax = ""; + simpleTax = ""; + + int seqTaxIndex = tax; + TaxNode seqTax = phyloTree->get(tax); + + while (seqTax.level != 0) { //while you are not at the root + + itBoot2 = confidenceScores.find(seqTaxIndex); //is this a classification we already have a count on + + int confidence = 0; + if (itBoot2 != confidenceScores.end()) { //already in confidence scores + confidence = confidenceScores[seqTaxIndex]; + } + + if (((confidence/(float)iters) * 100) >= confidenceThreshold) { + confidenceTax = seqTax.name + "(" + toString(((confidence/(float)iters) * 100)) + ");" + confidenceTax; + simpleTax = seqTax.name + ";" + simpleTax; + } + + seqTaxIndex = seqTax.parent; + seqTax = phyloTree->get(seqTax.parent); + } + + if (confidenceTax == "") { confidenceTax = "unclassified;"; simpleTax = "unclassified;"; } + return confidenceTax; + + } + catch(exception& e) { + m->errorOut(e, "Bayesian", "bootstrapResults"); + exit(1); + } +} +/**************************************************************************************************/ +int Bayesian::getMostProbableTaxonomy(vector queryKmer) { + try { + int indexofGenus = 0; + + double maxProbability = -1000000.0; + //find taxonomy with highest probability that this sequence is from it + for (int k = 0; k < genusNodes.size(); k++) { + //for each taxonomy calc its probability + double prob = 1.0; + for (int i = 0; i < queryKmer.size(); i++) { + prob += wordGenusProb[queryKmer[i]][k]; + } + + //is this the taxonomy with the greatest probability? + if (prob > maxProbability) { + indexofGenus = genusNodes[k]; + maxProbability = prob; + } + } + + return indexofGenus; + } + catch(exception& e) { + m->errorOut(e, "Bayesian", "getMostProbableTaxonomy"); + exit(1); + } +} +/************************************************************************************************* +map Bayesian::parseTaxMap(string newTax) { + try{ + + map parsed; + + newTax = newTax.substr(0, newTax.length()-1); //get rid of last ';' + + //parse taxonomy + string individual; + while (newTax.find_first_of(';') != -1) { + individual = newTax.substr(0,newTax.find_first_of(';')); + newTax = newTax.substr(newTax.find_first_of(';')+1, newTax.length()); + parsed[individual] = 1; + } + + //get last one + parsed[newTax] = 1; + + return parsed; + + } + catch(exception& e) { + m->errorOut(e, "Bayesian", "parseTax"); + exit(1); + } +} +/**************************************************************************************************/ +void Bayesian::readProbFile(ifstream& in, ifstream& inNum, string inName, string inNumName) { + try{ + + #ifdef USE_MPI + + int pid, num, num2, processors; + vector positions; + vector positions2; + + MPI_Status status; + MPI_File inMPI; + MPI_File inMPI2; + MPI_Comm_rank(MPI_COMM_WORLD, &pid); //find out who we are + MPI_Comm_size(MPI_COMM_WORLD, &processors); + int tag = 2001; + + char inFileName[1024]; + strcpy(inFileName, inNumName.c_str()); + + char inFileName2[1024]; + strcpy(inFileName2, inName.c_str()); + + MPI_File_open(MPI_COMM_WORLD, inFileName, MPI_MODE_RDONLY, MPI_INFO_NULL, &inMPI); //comm, filename, mode, info, filepointer + MPI_File_open(MPI_COMM_WORLD, inFileName2, MPI_MODE_RDONLY, MPI_INFO_NULL, &inMPI2); //comm, filename, mode, info, filepointer + + if (pid == 0) { + positions = setFilePosEachLine(inNumName, num); + positions2 = setFilePosEachLine(inName, num2); + + for(int i = 1; i < processors; i++) { + MPI_Send(&num, 1, MPI_INT, i, tag, MPI_COMM_WORLD); + MPI_Send(&positions[0], (num+1), MPI_LONG, i, tag, MPI_COMM_WORLD); + + MPI_Send(&num2, 1, MPI_INT, i, tag, MPI_COMM_WORLD); + MPI_Send(&positions2[0], (num2+1), MPI_LONG, i, tag, MPI_COMM_WORLD); + } + + }else{ + MPI_Recv(&num, 1, MPI_INT, 0, tag, MPI_COMM_WORLD, &status); + positions.resize(num+1); + MPI_Recv(&positions[0], (num+1), MPI_LONG, 0, tag, MPI_COMM_WORLD, &status); + + MPI_Recv(&num2, 1, MPI_INT, 0, tag, MPI_COMM_WORLD, &status); + positions2.resize(num2+1); + MPI_Recv(&positions2[0], (num2+1), MPI_LONG, 0, tag, MPI_COMM_WORLD, &status); + } + + //read numKmers + int length = positions2[1] - positions2[0]; + char* buf = new char[length]; + + MPI_File_read_at(inMPI2, positions2[0], buf, length, MPI_CHAR, &status); + + string tempBuf = buf; + if (tempBuf.length() > length) { tempBuf = tempBuf.substr(0, length); } + delete buf; + + istringstream iss (tempBuf,istringstream::in); + iss >> numKmers; + + //initialze probabilities + wordGenusProb.resize(numKmers); + + for (int j = 0; j < wordGenusProb.size(); j++) { wordGenusProb[j].resize(genusNodes.size()); } + + int kmer, name; + vector numbers; numbers.resize(numKmers); + float prob; + vector zeroCountProb; zeroCountProb.resize(numKmers); + + //read file + for(int i=0;i length) { tempBuf = tempBuf.substr(0, length); } + delete buf4; + + istringstream iss (tempBuf,istringstream::in); + iss >> zeroCountProb[i] >> numbers[i]; + } + + MPI_File_close(&inMPI); + + for(int i=1;i length) { tempBuf = tempBuf.substr(0, length); } + delete buf4; + + istringstream iss (tempBuf,istringstream::in); + + iss >> kmer; + + //set them all to zero value + for (int i = 0; i < genusNodes.size(); i++) { + wordGenusProb[kmer][i] = log(zeroCountProb[kmer] / (float) (genusTotals[i]+1)); + } + + //get probs for nonzero values + for (int i = 0; i < numbers[kmer]; i++) { + iss >> name >> prob; + wordGenusProb[kmer][name] = prob; + } + + } + MPI_File_close(&inMPI2); + MPI_Barrier(MPI_COMM_WORLD); //make everyone wait - just in case + #else + + in >> numKmers; gobble(in); + + //initialze probabilities + wordGenusProb.resize(numKmers); + + for (int j = 0; j < wordGenusProb.size(); j++) { wordGenusProb[j].resize(genusNodes.size()); } + + int kmer, name, count; count = 0; + vector num; num.resize(numKmers); + float prob; + vector zeroCountProb; zeroCountProb.resize(numKmers); + + while (inNum) { + inNum >> zeroCountProb[count] >> num[count]; + count++; + gobble(inNum); + } + inNum.close(); + + while(in) { + in >> kmer; + + //set them all to zero value + for (int i = 0; i < genusNodes.size(); i++) { + wordGenusProb[kmer][i] = log(zeroCountProb[kmer] / (float) (genusTotals[i]+1)); + } + + //get probs for nonzero values + for (int i = 0; i < num[kmer]; i++) { + in >> name >> prob; + wordGenusProb[kmer][name] = prob; + } + + gobble(in); + } + in.close(); + + #endif + } + catch(exception& e) { + m->errorOut(e, "Bayesian", "readProbFile"); exit(1); } } /**************************************************************************************************/ + + + + +