+//
+// kmerTree.cpp
+// pdsBayesian
+//
+// Created by Patrick Schloss on 4/3/12.
+// Copyright (c) 2012 University of Michigan. All rights reserved.
+//
+
+#include "kmernode.h"
+#include "kmertree.h"
+
+/**************************************************************************************************/
+
+KmerTree::KmerTree(string referenceFileName, string taxonomyFileName, int k, int cutoff) : Classify(), confidenceThreshold(cutoff), kmerSize(k){
+ try {
+ KmerNode* newNode = new KmerNode("Root", 0, kmerSize);
+ tree.push_back(newNode); // the tree is stored as a vector of elements of type TaxonomyNode
+
+ int power4s[14] = { 1, 4, 16, 64, 256, 1024, 4096, 16384, 65536, 262144, 1048576, 4194304, 16777216, 67108864 };
+ numPossibleKmers = power4s[kmerSize];
+
+ string refTaxonomy;
+
+ readTaxonomy(taxonomyFileName);
+
+ ifstream referenceFile;
+ m->openInputFile(referenceFileName, referenceFile);
+ bool error = false;
+ while(!referenceFile.eof()){
+
+ if (m->control_pressed) { break; }
+
+ Sequence seq(referenceFile); m->gobble(referenceFile);
+
+ if (seq.getName() != "") {
+ map<string, string>::iterator it = taxonomy.find(seq.getName());
+
+ if (it != taxonomy.end()) {
+ refTaxonomy = it->second; // lookup the taxonomy string for the current reference sequence
+ vector<int> kmerProfile = ripKmerProfile(seq.getUnaligned()); //convert to kmer vector
+ addTaxonomyToTree(seq.getName(), refTaxonomy, kmerProfile);
+ }else {
+ m->mothurOut(seq.getName() + " is in your reference file, but not in your taxonomy file, please correct.\n"); error = true;
+ }
+ }
+ }
+ referenceFile.close();
+
+ if (error) { m->control_pressed = true; }
+
+ numTaxa = (int)tree.size();
+ numLevels = 0;
+ for(int i=0;i<numTaxa;i++){
+ int level = tree[i]->getLevel();
+ if(level > numLevels){ numLevels = level; }
+ }
+ numLevels++;
+
+ aggregateThetas();
+
+ int dbSize = tree[0]->getNumSeqs();
+
+ for(int i=0;i<numTaxa;i++){
+ tree[i]->checkTheta();
+ tree[i]->setNumUniqueKmers(tree[0]->getNumUniqueKmers());
+ tree[i]->setTotalSeqs(dbSize);
+ }
+ }
+ catch(exception& e) {
+ m->errorOut(e, "KmerTree", "KmerTree");
+ exit(1);
+ }
+}
+
+/**************************************************************************************************/
+
+KmerTree::~KmerTree(){
+
+ for(int i=0;i<tree.size();i++){
+ delete tree[i];
+ }
+
+}
+/**********************************************************************************************************************/
+
+vector<int> KmerTree::ripKmerProfile(string sequence){
+ try {
+ // assume all input sequences are unaligned
+
+ int power4s[14] = { 1, 4, 16, 64, 256, 1024, 4096, 16384, 65536, 262144, 1048576, 4194304, 16777216, 67108864 };
+
+ int nKmers = (int)sequence.length() - kmerSize + 1;
+
+ vector<int> kmerProfile(numPossibleKmers + 1, 0);
+
+ for(int i=0;i<nKmers;i++){
+
+ if (m->control_pressed) { break; }
+
+ int kmer = 0;
+ for(int j=0;j<kmerSize;j++){
+ if(toupper(sequence[j+i]) == 'A') { kmer += (0 * power4s[kmerSize-j-1]); }
+ else if(toupper(sequence[j+i]) == 'C') { kmer += (1 * power4s[kmerSize-j-1]); }
+ else if(toupper(sequence[j+i]) == 'G') { kmer += (2 * power4s[kmerSize-j-1]); }
+ else if(toupper(sequence[j+i]) == 'U') { kmer += (3 * power4s[kmerSize-j-1]); }
+ else if(toupper(sequence[j+i]) == 'T') { kmer += (3 * power4s[kmerSize-j-1]); }
+ else { kmer = power4s[kmerSize]; j = kmerSize; }
+ }
+ kmerProfile[kmer] = 1;
+ }
+
+ return kmerProfile;
+ }
+ catch(exception& e) {
+ m->errorOut(e, "KmerTree", "ripKmerProfile");
+ exit(1);
+ }
+}
+
+/**************************************************************************************************/
+
+int KmerTree::addTaxonomyToTree(string seqName, string taxonomy, vector<int>& sequence){
+ try {
+ KmerNode* newNode;
+ string taxonName = "";
+ int treePosition = 0; // the root is element 0
+
+
+ int level = 1;
+
+ for(int i=0;i<taxonomy.length();i++){ // step through taxonomy string...
+
+ if (m->control_pressed) { break; }
+ if(taxonomy[i] == ';'){ // looking for semicolons...
+
+ if (taxonName == "") { m->mothurOut(seqName + " has an error in the taxonomy. This may be due to a ;;"); m->mothurOutEndLine(); m->control_pressed = true; }
+
+ int newIndex = tree[treePosition]->getChildIndex(taxonName);// look to see if your current node already
+ // has a child with the new taxonName
+ if(newIndex != -1) { treePosition = newIndex; } // if you've seen it before, jump to that
+ else { // position in the tree
+ int newChildIndex = (int)tree.size(); // otherwise, we'll have to create one...
+ tree[treePosition]->makeChild(taxonName, newChildIndex);
+
+ newNode = new KmerNode(taxonName, level, kmerSize);
+
+ newNode->setParent(treePosition);
+
+ tree.push_back(newNode);
+ treePosition = newChildIndex;
+ }
+
+ // sequence data to that node to update that node's theta - seems slow...
+ taxonName = ""; // clear out the taxon name that we will build as we look
+ level++;
+
+ } // for a semicolon
+ else{
+ taxonName += taxonomy[i]; // keep adding letters until we reach a semicolon
+ }
+ }
+
+ tree[treePosition]->loadSequence(sequence); // now that we've gotten to the correct node, add the
+
+ return 0;
+ }
+ catch(exception& e) {
+ m->errorOut(e, "KmerTree", "addTaxonomyToTree");
+ exit(1);
+ }
+
+}
+
+/**************************************************************************************************/
+
+int KmerTree::aggregateThetas(){
+ try {
+ vector<vector<int> > levelMatrix(numLevels+1);
+
+ for(int i=0;i<tree.size();i++){
+ if (m->control_pressed) { return 0; }
+ levelMatrix[tree[i]->getLevel()].push_back(i);
+ }
+
+ for(int i=numLevels-1;i>0;i--) {
+ if (m->control_pressed) { return 0; }
+
+ for(int j=0;j<levelMatrix[i].size();j++){
+
+ KmerNode* holder = tree[levelMatrix[i][j]];
+
+ tree[holder->getParent()]->addThetas(holder->getTheta(), holder->getNumSeqs());
+ }
+ }
+
+ return 0;
+ }
+ catch(exception& e) {
+ m->errorOut(e, "KmerTree", "aggregateThetas");
+ exit(1);
+ }
+}
+
+/**************************************************************************************************/
+
+int KmerTree::getMinRiskIndexKmer(vector<int>& sequence, vector<int>& taxaIndices, vector<double>& probabilities){
+ try {
+ int numProbs = (int)probabilities.size();
+
+ vector<double> G(numProbs, 0.2); //a random sequence will, on average, be 20% similar to any other sequence; not sure that this holds up for kmers; whatever.
+ vector<double> risk(numProbs, 0);
+
+ for(int i=1;i<numProbs;i++){ //use if you want the outlier group
+ if (m->control_pressed) { return 0; }
+ G[i] = tree[taxaIndices[i]]->getSimToConsensus(sequence);
+ }
+
+ double minRisk = 1e6;
+ int minRiskIndex = 0;
+
+ for(int i=0;i<numProbs;i++){
+ if (m->control_pressed) { return 0; }
+ for(int j=0;j<numProbs;j++){
+ if(i != j){
+ risk[i] += probabilities[j] * G[j];
+ }
+ }
+
+ if(risk[i] < minRisk){
+ minRisk = risk[i];
+ minRiskIndex = i;
+ }
+ }
+
+ return minRiskIndex;
+ }
+ catch(exception& e) {
+ m->errorOut(e, "KmerTree", "getMinRiskIndexKmer");
+ exit(1);
+ }
+}
+
+/**************************************************************************************************/
+
+int KmerTree::sanityCheck(vector<vector<int> >& indices, vector<int>& maxIndices){
+ try {
+ int finalLevel = (int)indices.size()-1;
+
+ for(int position=1;position<indices.size();position++){
+ if (m->control_pressed) { return 0; }
+ int predictedParent = tree[indices[position][maxIndices[position]]]->getParent();
+ int actualParent = indices[position-1][maxIndices[position-1]];
+
+ if(predictedParent != actualParent){
+ finalLevel = position - 1;
+ return finalLevel;
+ }
+ }
+ return finalLevel;
+ }
+ catch(exception& e) {
+ m->errorOut(e, "KmerTree", "sanityCheck");
+ exit(1);
+ }
+}
+
+/**************************************************************************************************/
+string KmerTree::getTaxonomy(Sequence* thisSeq){
+ try {
+ string seqName = thisSeq->getName(); string querySequence = thisSeq->getAligned(); string taxonProbabilityString = "";
+ string unalignedSeq = thisSeq->getUnaligned();
+
+ double logPOutlier = (querySequence.length() - kmerSize + 1) * log(1.0/(double)tree[0]->getNumUniqueKmers());
+
+ vector<int> queryProfile = ripKmerProfile(unalignedSeq); //convert to kmer vector
+
+ vector<vector<double> > pXgivenKj_D_j(numLevels);
+ vector<vector<int> > indices(numLevels);
+ for(int i=0;i<numLevels;i++){
+ if (m->control_pressed) { return taxonProbabilityString; }
+ pXgivenKj_D_j[i].push_back(logPOutlier);
+ indices[i].push_back(-1);
+ }
+
+ for(int i=0;i<numTaxa;i++){
+ if (m->control_pressed) { return taxonProbabilityString; }
+ pXgivenKj_D_j[tree[i]->getLevel()].push_back(tree[i]->getPxGivenkj_D_j(queryProfile));
+ indices[tree[i]->getLevel()].push_back(i);
+ }
+
+ vector<double> sumLikelihood(numLevels, 0);
+ vector<double> bestPosterior(numLevels, 0);
+ vector<int> maxIndex(numLevels, 0);
+ int maxPosteriorIndex;
+
+ //let's find the best level and taxa within that level
+ for(int i=0;i<numLevels;i++){ //go across all j's - from the root to genus
+ if (m->control_pressed) { return taxonProbabilityString; }
+
+ int numTaxaInLevel = (int)indices[i].size();
+
+ vector<double> posteriors(numTaxaInLevel, 0);
+ sumLikelihood[i] = getLogExpSum(pXgivenKj_D_j[i], maxPosteriorIndex);
+
+ maxPosteriorIndex = 0;
+ for(int j=0;j<numTaxaInLevel;j++){
+ posteriors[j] = exp(pXgivenKj_D_j[i][j] - sumLikelihood[i]);
+ if(posteriors[j] > posteriors[maxPosteriorIndex]){
+ maxPosteriorIndex = j;
+ }
+
+ }
+
+ maxIndex[i] = getMinRiskIndexKmer(queryProfile, indices[i], posteriors);
+
+ maxIndex[i] = maxPosteriorIndex;
+ bestPosterior[i] = posteriors[maxIndex[i]];
+ }
+
+ // vector<double> pX_level(numLevels, 0);
+ //
+ // for(int i=0;i<numLevels;i++){
+ // pX_level[i] = pXgivenKj_D_j[i][maxIndex[i]] - tree[indices[i][maxIndex[i]]]->getNumSeqs();
+ // }
+ //
+ // int max_pLevel_X_index = -1;
+ // double pX_level_sum = getLogExpSum(pX_level, max_pLevel_X_index);
+ // double max_pLevel_X = exp(pX_level[max_pLevel_X_index] - pX_level_sum);
+ //
+ // vector<double> pLevel_X(numLevels, 0);
+ // for(int i=0;i<numLevels;i++){
+ // pLevel_X[i] = exp(pX_level[i] - pX_level_sum);
+ // }
+
+ int saneDepth = sanityCheck(indices, maxIndex);
+
+
+ // stringstream levelProbabilityOutput;
+ // levelProbabilityOutput.setf(ios::fixed, ios::floatfield);
+ // levelProbabilityOutput.setf(ios::showpoint);
+
+
+ //taxonProbabilityOutput << seqName << '\t';
+ // taxonProbabilityOutput << seqName << '(' << max_pLevel_X_index << ';' << max_pLevel_X << ')' << '\t';
+ // levelProbabilityOutput << seqName << '(' << max_pLevel_X_index << ';' << max_pLevel_X << ')' << '\t';
+ simpleTax = "";
+ int savedspot = 1;
+ taxonProbabilityString = "";
+ for(int i=1;i<=saneDepth;i++){
+ if (m->control_pressed) { return taxonProbabilityString; }
+ int confidenceScore = (int) (bestPosterior[i] * 100);
+ if (confidenceScore >= confidenceThreshold) {
+ if(indices[i][maxIndex[i]] != -1){
+ taxonProbabilityString += tree[indices[i][maxIndex[i]]]->getName() + "(" + toString(confidenceScore) + ");";
+ simpleTax += tree[indices[i][maxIndex[i]]]->getName() + ";";
+
+ // levelProbabilityOutput << tree[indices[i][maxIndex[i]]]->getName() << '(' << setprecision(6) << pLevel_X[i] << ");";
+ }
+ else{
+ taxonProbabilityString += "unclassified" + '(' + toString(confidenceScore) + ");";
+ // levelProbabilityOutput << "unclassified" << '(' << setprecision(6) << pLevel_X[i] << ");";
+ simpleTax += "unclassified;";
+ }
+ }else { break; }
+ savedspot = i;
+ }
+
+
+
+ for(int i=savedspot+1;i<numLevels;i++){
+ if (m->control_pressed) { return taxonProbabilityString; }
+ taxonProbabilityString += "unclassified(0);";
+ simpleTax += "unclassified;";
+ }
+
+ return taxonProbabilityString;
+ }
+ catch(exception& e) {
+ m->errorOut(e, "KmerTree", "getTaxonomy");
+ exit(1);
+ }
+}
+
+
+/**************************************************************************************************/
+