]> git.donarmstrong.com Git - mothur.git/blobdiff - classifyseqscommand.cpp
1.12.1
[mothur.git] / classifyseqscommand.cpp
index e4fcb2b9b61c926dfe996d5c643934cc9a926739..a95f436dd7ffdf4403d316be77a1f8ab580f21f5 100644 (file)
@@ -11,6 +11,7 @@
 #include "sequence.hpp"
 #include "bayesian.h"
 #include "phylotree.h"
+#include "phylosummary.h"
 #include "knn.h"
 
 //**********************************************************************************************************************
@@ -25,13 +26,13 @@ ClassifySeqsCommand::ClassifySeqsCommand(string option)  {
                else {
                        
                        //valid paramters for this command
-                       string AlignArray[] =  {"template","fasta","name","search","ksize","method","processors","taxonomy","match","mismatch","gapopen","gapextend","numwanted","cutoff","probs","iters", "outputdir","inputdir"};
+                       string AlignArray[] =  {"template","fasta","name","group","search","ksize","method","processors","taxonomy","match","mismatch","gapopen","gapextend","numwanted","cutoff","probs","iters", "outputdir","inputdir"};
                        vector<string> myArray (AlignArray, AlignArray+(sizeof(AlignArray)/sizeof(string)));
                        
                        OptionParser parser(option);
                        map<string, string> parameters = parser.getParameters(); 
                        
-                       ValidParameters validParameter;
+                       ValidParameters validParameter("classify.seqs");
                        map<string, string>::iterator it;
                        
                        //check to make sure all parameters are valid for command
@@ -62,6 +63,14 @@ ClassifySeqsCommand::ClassifySeqsCommand(string option)  {
                                        //if the user has not given a path then, add inputdir. else leave path alone.
                                        if (path == "") {       parameters["taxonomy"] = inputDir + it->second;         }
                                }
+                               
+                               it = parameters.find("group");
+                               //user has given a template file
+                               if(it != parameters.end()){ 
+                                       path = hasPath(it->second);
+                                       //if the user has not given a path then, add inputdir. else leave path alone.
+                                       if (path == "") {       parameters["group"] = inputDir + it->second;            }
+                               }
                        }
 
                        //check for required parameters
@@ -73,6 +82,7 @@ ClassifySeqsCommand::ClassifySeqsCommand(string option)  {
                        }
                        else if (templateFileName == "not open") { abort = true; }      
                        
+                                               
                        fastaFileName = validParameter.validFile(parameters, "fasta", false);
                        if (fastaFileName == "not found") { m->mothurOut("fasta is a required parameter for the classify.seqs command."); m->mothurOutEndLine(); abort = true;  }
                        else { 
@@ -97,7 +107,17 @@ ClassifySeqsCommand::ClassifySeqsCommand(string option)  {
                                        #endif
                                        
                                        ifstream in;
-                                       ableToOpen = openInputFile(fastaFileNames[i], in);
+                                       ableToOpen = openInputFile(fastaFileNames[i], in, "noerror");
+                               
+                                       //if you can't open it, try default location
+                                       if (ableToOpen == 1) {
+                                               if (m->getDefaultPath() != "") { //default path is set
+                                                       string tryPath = m->getDefaultPath() + getSimpleName(fastaFileNames[i]);
+                                                       m->mothurOut("Unable to open " + fastaFileNames[i] + ". Trying default " + tryPath); m->mothurOutEndLine();
+                                                       ableToOpen = openInputFile(tryPath, in, "noerror");
+                                                       fastaFileNames[i] = tryPath;
+                                               }
+                                       }
                                        in.close();
                                        
                                        #ifdef USE_MPI  
@@ -112,7 +132,7 @@ ClassifySeqsCommand::ClassifySeqsCommand(string option)  {
                                        #endif
                                        
                                        if (ableToOpen == 1) { 
-                                               m->mothurOut(fastaFileNames[i] + " will be disregarded."); m->mothurOutEndLine(); 
+                                               m->mothurOut("Unable to open " + fastaFileNames[i] + ". It will be disregarded."); m->mothurOutEndLine(); 
                                                //erase from file list
                                                fastaFileNames.erase(fastaFileNames.begin()+i);
                                                i--;
@@ -158,7 +178,17 @@ ClassifySeqsCommand::ClassifySeqsCommand(string option)  {
                                        #endif
 
                                        ifstream in;
-                                       ableToOpen = openInputFile(namefileNames[i], in);
+                                       ableToOpen = openInputFile(namefileNames[i], in, "noerror");
+                               
+                                       //if you can't open it, try default location
+                                       if (ableToOpen == 1) {
+                                               if (m->getDefaultPath() != "") { //default path is set
+                                                       string tryPath = m->getDefaultPath() + getSimpleName(namefileNames[i]);
+                                                       m->mothurOut("Unable to open " + namefileNames[i] + ". Trying default " + tryPath); m->mothurOutEndLine();
+                                                       ableToOpen = openInputFile(tryPath, in, "noerror");
+                                                       namefileNames[i] = tryPath;
+                                               }
+                                       }
                                        in.close();
                                        
                                        #ifdef USE_MPI  
@@ -171,8 +201,14 @@ ClassifySeqsCommand::ClassifySeqsCommand(string option)  {
                                                }
                                                
                                        #endif
-                                       if (ableToOpen == 1) {  m->mothurOut("Unable to match name file with fasta file."); m->mothurOutEndLine(); abort = true;        }
                                        
+                                       if (ableToOpen == 1) { 
+                                               m->mothurOut("Unable to open " + namefileNames[i] + ". It will be disregarded."); m->mothurOutEndLine();  abort = true;
+                                               //erase from file list
+                                               namefileNames.erase(namefileNames.begin()+i);
+                                               i--;
+                                       }
+
                                }
                        }
 
@@ -180,6 +216,68 @@ ClassifySeqsCommand::ClassifySeqsCommand(string option)  {
                                if (namefileNames.size() != fastaFileNames.size()) { abort = true; m->mothurOut("If you provide a name file, you must have one for each fasta file."); m->mothurOutEndLine(); }
                        }
                        
+                       groupfile = validParameter.validFile(parameters, "group", false);
+                       if (groupfile == "not found") { groupfile = "";  }
+                       else { 
+                               splitAtDash(groupfile, groupfileNames);
+                               
+                               //go through files and make sure they are good, if not, then disregard them
+                               for (int i = 0; i < groupfileNames.size(); i++) {
+                                       if (inputDir != "") {
+                                               string path = hasPath(groupfileNames[i]);
+                                               //if the user has not given a path then, add inputdir. else leave path alone.
+                                               if (path == "") {       groupfileNames[i] = inputDir + groupfileNames[i];               }
+                                       }
+                                       int ableToOpen;
+                                       
+                                       #ifdef USE_MPI  
+                                               int pid;
+                                               MPI_Comm_size(MPI_COMM_WORLD, &processors); //set processors to the number of mpi processes running
+                                               MPI_Comm_rank(MPI_COMM_WORLD, &pid); //find out who we are
+                               
+                                               if (pid == 0) {
+                                       #endif
+
+                                       ifstream in;
+                                       ableToOpen = openInputFile(groupfileNames[i], in, "noerror");
+                               
+                                       //if you can't open it, try default location
+                                       if (ableToOpen == 1) {
+                                               if (m->getDefaultPath() != "") { //default path is set
+                                                       string tryPath = m->getDefaultPath() + getSimpleName(groupfileNames[i]);
+                                                       m->mothurOut("Unable to open " + groupfileNames[i] + ". Trying default " + tryPath); m->mothurOutEndLine();
+                                                       ableToOpen = openInputFile(tryPath, in, "noerror");
+                                                       groupfileNames[i] = tryPath;
+                                               }
+                                       }
+                                       in.close();
+                                       
+                                       #ifdef USE_MPI  
+                                                       for (int j = 1; j < processors; j++) {
+                                                               MPI_Send(&ableToOpen, 1, MPI_INT, j, 2001, MPI_COMM_WORLD); 
+                                                       }
+                                               }else{
+                                                       MPI_Status status;
+                                                       MPI_Recv(&ableToOpen, 1, MPI_INT, 0, 2001, MPI_COMM_WORLD, &status);
+                                               }
+                                               
+                                       #endif
+                                       
+                                       if (ableToOpen == 1) { 
+                                               m->mothurOut("Unable to open " + groupfileNames[i] + ". It will be disregarded."); m->mothurOutEndLine(); groupfileNames[i] = "";
+                                               //erase from file list
+                                               groupfileNames.erase(groupfileNames.begin()+i);
+                                               i--;
+                                       }
+                               }
+                       }
+
+                       if (groupfile != "") {
+                               if (groupfileNames.size() != fastaFileNames.size()) { abort = true; m->mothurOut("If you provide a group file, you must have one for each fasta file."); m->mothurOutEndLine(); }
+                       }else {
+                               for (int i = 0; i < fastaFileNames.size(); i++) {  groupfileNames.push_back("");  }
+                       }
+                       
                        //check for optional parameter and set defaults
                        // ...at some point should added some additional type checking...
                        string temp;
@@ -250,6 +348,7 @@ void ClassifySeqsCommand::help(){
                m->mothurOut("The template, fasta and taxonomy parameters are required. You may enter multiple fasta files by separating their names with dashes. ie. fasta=abrecovery.fasta-amzon.fasta \n");
                m->mothurOut("The search parameter allows you to specify the method to find most similar template.  Your options are: suffix, kmer, blast and distance. The default is kmer.\n");
                m->mothurOut("The name parameter allows you add a names file with your fasta file, if you enter multiple fasta files, you must enter matching names files for them.\n");
+               m->mothurOut("The group parameter allows you add a group file so you can have the summary totals broken up by group.\n");
                m->mothurOut("The method parameter allows you to specify classification method to use.  Your options are: bayesian and knn. The default is bayesian.\n");
                m->mothurOut("The ksize parameter allows you to specify the kmer size for finding most similar template to candidate.  The default is 8.\n");
                m->mothurOut("The processors parameter allows you to specify the number of processors to use. The default is 1.\n");
@@ -262,7 +361,7 @@ void ClassifySeqsCommand::help(){
                m->mothurOut("The gapextend parameter allows you to specify the penalty for extending a gap in an alignment.  The default is -1.0.\n");
                m->mothurOut("The numwanted parameter allows you to specify the number of sequence matches you want with the knn method.  The default is 10.\n");
                m->mothurOut("The cutoff parameter allows you to specify a bootstrap confidence threshold for your taxonomy.  The default is 0.\n");
-               m->mothurOut("The probs parameter shut off the bootstrapping results for the bayesian method. The default is true, meaning you want the bootstrapping to be run.\n");
+               m->mothurOut("The probs parameter shuts off the bootstrapping results for the bayesian method. The default is true, meaning you want the bootstrapping to be shown.\n");
                m->mothurOut("The iters parameter allows you to specify how many iterations to do when calculating the bootstrap confidence score for your taxonomy with the bayesian method.  The default is 100.\n");
                m->mothurOut("The classify.seqs command should be in the following format: \n");
                m->mothurOut("classify.seqs(template=yourTemplateFile, fasta=yourFastaFile, method=yourClassificationMethod, search=yourSearchmethod, ksize=yourKmerSize, taxonomy=yourTaxonomyFile, processors=yourProcessors) \n");
@@ -300,10 +399,15 @@ int ClassifySeqsCommand::execute(){
                
                        m->mothurOut("Classifying sequences from " + fastaFileNames[s] + " ..." ); m->mothurOutEndLine();
                        
+                       string RippedTaxName = getRootName(getSimpleName(taxonomyFileName));
+                       RippedTaxName = getExtension(RippedTaxName.substr(0, RippedTaxName.length()-1));
+                       if (RippedTaxName[0] == '.') { RippedTaxName = RippedTaxName.substr(1, RippedTaxName.length()); }
+                       RippedTaxName +=  "."; 
+               
                        if (outputDir == "") { outputDir += hasPath(fastaFileNames[s]); }
-                       string newTaxonomyFile = outputDir + getRootName(getSimpleName(fastaFileNames[s])) + getRootName(getSimpleName(taxonomyFileName)) + "taxonomy";
+                       string newTaxonomyFile = outputDir + getRootName(getSimpleName(fastaFileNames[s])) + RippedTaxName + "taxonomy";
                        string tempTaxonomyFile = outputDir + getRootName(getSimpleName(fastaFileNames[s])) + "taxonomy.temp";
-                       string taxSummary = outputDir + getRootName(getSimpleName(fastaFileNames[s])) + getRootName(getSimpleName(taxonomyFileName)) + "tax.summary";
+                       string taxSummary = outputDir + getRootName(getSimpleName(fastaFileNames[s])) + RippedTaxName + "tax.summary";
                        
                        outputNames.push_back(newTaxonomyFile);
                        outputNames.push_back(taxSummary);
@@ -315,7 +419,7 @@ int ClassifySeqsCommand::execute(){
 #ifdef USE_MPI 
                                int pid, end, numSeqsPerProcessor; 
                                int tag = 2001;
-                               vector<long> MPIPos;
+                               vector<unsigned long int> MPIPos;
                                
                                MPI_Status status; 
                                MPI_Comm_rank(MPI_COMM_WORLD, &pid); //find out who we are
@@ -328,39 +432,49 @@ int ClassifySeqsCommand::execute(){
                                int outMode=MPI_MODE_CREATE|MPI_MODE_WRONLY; 
                                int inMode=MPI_MODE_RDONLY; 
                                
-                               char* outNewTax = new char[newTaxonomyFile.length()];\r
-                               memcpy(outNewTax, newTaxonomyFile.c_str(), newTaxonomyFile.length());
-                       
-                               char* outTempTax = new char[tempTaxonomyFile.length()];\r
-                               memcpy(outTempTax, tempTaxonomyFile.c_str(), tempTaxonomyFile.length());
+                               //char* outNewTax = new char[newTaxonomyFile.length()];
+                               //memcpy(outNewTax, newTaxonomyFile.c_str(), newTaxonomyFile.length());
+                               
+                               char outNewTax[1024];
+                               strcpy(outNewTax, newTaxonomyFile.c_str());
+
+                               //char* outTempTax = new char[tempTaxonomyFile.length()];
+                               //memcpy(outTempTax, tempTaxonomyFile.c_str(), tempTaxonomyFile.length());
+                               
+                               char outTempTax[1024];
+                               strcpy(outTempTax, tempTaxonomyFile.c_str());
 
-                               char* inFileName = new char[fastaFileNames[s].length()];\r
-                               memcpy(inFileName, fastaFileNames[s].c_str(), fastaFileNames[s].length());
+                               //char* inFileName = new char[fastaFileNames[s].length()];
+                               //memcpy(inFileName, fastaFileNames[s].c_str(), fastaFileNames[s].length());
+                               
+                               char inFileName[1024];
+                               strcpy(inFileName, fastaFileNames[s].c_str());
 
                                MPI_File_open(MPI_COMM_WORLD, inFileName, inMode, MPI_INFO_NULL, &inMPI);  //comm, filename, mode, info, filepointer
                                MPI_File_open(MPI_COMM_WORLD, outNewTax, outMode, MPI_INFO_NULL, &outMPINewTax);
                                MPI_File_open(MPI_COMM_WORLD, outTempTax, outMode, MPI_INFO_NULL, &outMPITempTax);
                                
-                               delete outNewTax;
-                               delete outTempTax;
-                               delete inFileName;
+                               //delete outNewTax;
+                               //delete outTempTax;
+                               //delete inFileName;
 
                                if (m->control_pressed) {  MPI_File_close(&inMPI);  MPI_File_close(&outMPINewTax);   MPI_File_close(&outMPITempTax);  delete classify; return 0;  }
-
-                               if(namefile != "") {  MPIReadNamesFile(namefileNames[s]);  }
                                
                                if (pid == 0) { //you are the root process 
                                        
                                        MPIPos = setFilePosFasta(fastaFileNames[s], numFastaSeqs); //fills MPIPos, returns numSeqs
                                        
                                        //send file positions to all processes
-                                       MPI_Bcast(&numFastaSeqs, 1, MPI_INT, 0, MPI_COMM_WORLD);  //send numSeqs
-                                       MPI_Bcast(&MPIPos[0], (numFastaSeqs+1), MPI_LONG, 0, MPI_COMM_WORLD); //send file pos   
+                                       for(int i = 1; i < processors; i++) { 
+                                               MPI_Send(&numFastaSeqs, 1, MPI_INT, i, tag, MPI_COMM_WORLD);
+                                               MPI_Send(&MPIPos[0], (numFastaSeqs+1), MPI_LONG, i, tag, MPI_COMM_WORLD);
+                                       }
                                        
                                        //figure out how many sequences you have to align
                                        numSeqsPerProcessor = numFastaSeqs / processors;
-                                       if(pid == (processors - 1)){    numSeqsPerProcessor = numFastaSeqs - pid * numSeqsPerProcessor;         }
                                        int startIndex =  pid * numSeqsPerProcessor;
+                                       if(pid == (processors - 1)){    numSeqsPerProcessor = numFastaSeqs - pid * numSeqsPerProcessor;         }
+                                       
                                
                                        //align your part
                                        driverMPI(startIndex, numSeqsPerProcessor, inMPI, outMPINewTax, outMPITempTax, MPIPos);
@@ -372,14 +486,15 @@ int ClassifySeqsCommand::execute(){
                                                MPI_Recv(&done, 1, MPI_INT, i, tag, MPI_COMM_WORLD, &status);
                                        }
                                }else{ //you are a child process
-                                       MPI_Bcast(&numFastaSeqs, 1, MPI_INT, 0, MPI_COMM_WORLD); //get numSeqs
+                                       MPI_Recv(&numFastaSeqs, 1, MPI_INT, 0, tag, MPI_COMM_WORLD, &status);
                                        MPIPos.resize(numFastaSeqs+1);
-                                       MPI_Bcast(&MPIPos[0], (numFastaSeqs+1), MPI_LONG, 0, MPI_COMM_WORLD); //get file positions
+                                       MPI_Recv(&MPIPos[0], (numFastaSeqs+1), MPI_LONG, 0, tag, MPI_COMM_WORLD, &status);
                                        
                                        //figure out how many sequences you have to align
                                        numSeqsPerProcessor = numFastaSeqs / processors;
-                                       if(pid == (processors - 1)){    numSeqsPerProcessor = numFastaSeqs - pid * numSeqsPerProcessor;         }
                                        int startIndex =  pid * numSeqsPerProcessor;
+                                       if(pid == (processors - 1)){    numSeqsPerProcessor = numFastaSeqs - pid * numSeqsPerProcessor;         }
+                                       
                                        
                                        //align your part
                                        driverMPI(startIndex, numSeqsPerProcessor, inMPI, outMPINewTax, outMPITempTax, MPIPos);
@@ -394,28 +509,14 @@ int ClassifySeqsCommand::execute(){
                                MPI_File_close(&inMPI);
                                MPI_File_close(&outMPINewTax);
                                MPI_File_close(&outMPITempTax);
+                               MPI_Barrier(MPI_COMM_WORLD); //make everyone wait - just in case
                                
 #else
-                       //read namefile
-                       if(namefile != "") {
-                               nameMap.clear(); //remove old names
-                               
-                               ifstream inNames;
-                               openInputFile(namefileNames[s], inNames);
-                               
-                               string firstCol, secondCol;
-                               while(!inNames.eof()) {
-                                       inNames >> firstCol >> secondCol; gobble(inNames);
-                                       nameMap[firstCol] = getNumNames(secondCol);  //ex. seq1 seq1,seq3,seq5 -> seq1 = 3.
-                               }
-                               inNames.close();
-                       }
-
-       #if defined (__APPLE__) || (__MACH__) || (linux) || (__linux)
+               #if defined (__APPLE__) || (__MACH__) || (linux) || (__linux)
                        if(processors == 1){
                                ifstream inFASTA;
                                openInputFile(fastaFileNames[s], inFASTA);
-                               numFastaSeqs=count(istreambuf_iterator<char>(inFASTA),istreambuf_iterator<char>(), '>');
+                               getNumSeqs(inFASTA, numFastaSeqs);
                                inFASTA.close();
                                
                                lines.push_back(new linePair(0, numFastaSeqs));
@@ -423,7 +524,7 @@ int ClassifySeqsCommand::execute(){
                                driver(lines[0], newTaxonomyFile, tempTaxonomyFile, fastaFileNames[s]);
                        }
                        else{
-                               vector<int> positions;
+                               vector<unsigned long int> positions;
                                processIDS.resize(0);
                                
                                ifstream inFASTA;
@@ -433,7 +534,7 @@ int ClassifySeqsCommand::execute(){
                                while(!inFASTA.eof()){
                                        input = getline(inFASTA);
                                        if (input.length() != 0) {
-                                               if(input[0] == '>'){    int pos = inFASTA.tellg(); positions.push_back(pos - input.length() - 1);       }
+                                               if(input[0] == '>'){    unsigned long int pos = inFASTA.tellg(); positions.push_back(pos - input.length() - 1); }
                                        }
                                }
                                inFASTA.close();
@@ -441,9 +542,9 @@ int ClassifySeqsCommand::execute(){
                                numFastaSeqs = positions.size();
                                
                                int numSeqsPerProcessor = numFastaSeqs / processors;
-                               
+       
                                for (int i = 0; i < processors; i++) {
-                                       int startPos = positions[ i * numSeqsPerProcessor ];
+                                       unsigned long int startPos = positions[ i * numSeqsPerProcessor ];
                                        if(i == processors - 1){
                                                numSeqsPerProcessor = numFastaSeqs - i * numSeqsPerProcessor;
                                        }
@@ -465,7 +566,7 @@ int ClassifySeqsCommand::execute(){
        #else
                        ifstream inFASTA;
                        openInputFile(fastaFileNames[s], inFASTA);
-                       numFastaSeqs=count(istreambuf_iterator<char>(inFASTA),istreambuf_iterator<char>(), '>');
+                       getNumSeqs(inFASTA, numFastaSeqs);
                        inFASTA.close();
                        
                        lines.push_back(new linePair(0, numFastaSeqs));
@@ -474,54 +575,81 @@ int ClassifySeqsCommand::execute(){
        #endif  
 #endif
 
+               m->mothurOutEndLine();
+               m->mothurOut("It took " + toString(time(NULL) - start) + " secs to classify " + toString(numFastaSeqs) + " sequences."); m->mothurOutEndLine(); m->mothurOutEndLine();
+               start = time(NULL);
+
+
                #ifdef USE_MPI  
                        if (pid == 0) {  //this part does not need to be paralellized
+                       
+                               if(namefile != "") { m->mothurOut("Reading " + namefileNames[s] + "..."); cout.flush();  MPIReadNamesFile(namefileNames[s]);  m->mothurOut("  Done."); m->mothurOutEndLine(); }
+               #else
+                       //read namefile
+                       if(namefile != "") {
+                       
+                           m->mothurOut("Reading " + namefileNames[s] + "..."); cout.flush();
+                               
+                               nameMap.clear(); //remove old names
+                               
+                               ifstream inNames;
+                               openInputFile(namefileNames[s], inNames);
+                               
+                               string firstCol, secondCol;
+                               while(!inNames.eof()) {
+                                       inNames >> firstCol >> secondCol; gobble(inNames);
+                                       
+                                       vector<string> temp;
+                                       splitAtComma(secondCol, temp);
+                       
+                                       nameMap[firstCol] = temp;  
+                               }
+                               inNames.close();
+                               
+                               m->mothurOut("  Done."); m->mothurOutEndLine();
+                       }
                #endif
 
-                       //make taxonomy tree from new taxonomy file 
-                       PhyloTree taxaBrowser;
+                       string group = "";
+                       if (groupfile != "") {  group = groupfileNames[s]; }
                        
-                       if (m->control_pressed) {  for (int i = 0; i < outputNames.size(); i++) {       remove(outputNames[i].c_str()); } delete classify; return 0; }
+                       PhyloSummary taxaSum(taxonomyFileName, group);
                        
-                       ifstream in;
-                       openInputFile(tempTaxonomyFile, in);
+                       if (m->control_pressed) {  for (int i = 0; i < outputNames.size(); i++) {       remove(outputNames[i].c_str()); } delete classify; return 0; }
                
-                       //read in users taxonomy file and add sequences to tree
-                       string name, taxon;
-                       while(!in.eof()){
-                               in >> name >> taxon; gobble(in);
+                       if (namefile == "") {  taxaSum.summarize(tempTaxonomyFile);  }
+                       else {
+                               ifstream in;
+                               openInputFile(tempTaxonomyFile, in);
                                
-                               if (m->control_pressed) {  for (int i = 0; i < outputNames.size(); i++) {       remove(outputNames[i].c_str()); } remove(tempTaxonomyFile.c_str()); delete classify; return 0; }
+                               //read in users taxonomy file and add sequences to tree
+                               string name, taxon;
                                
-                               if (namefile != "") {
+                               while(!in.eof()){
+                                       in >> name >> taxon; gobble(in);
+                                       
                                        itNames = nameMap.find(name);
                
                                        if (itNames == nameMap.end()) { 
                                                m->mothurOut(name + " is not in your name file please correct."); m->mothurOutEndLine(); exit(1);
                                        }else{
-                                               for (int i = 0; i < itNames->second; i++) { 
-                                                       taxaBrowser.addSeqToTree(name+toString(i), taxon);  //add it as many times as there are identical seqs
+                                               for (int i = 0; i < itNames->second.size(); i++) { 
+                                                       taxaSum.addSeqToTree(itNames->second[i], taxon);  //add it as many times as there are identical seqs
                                                }
+                                               itNames->second.clear();
+                                               nameMap.erase(itNames->first);
                                        }
-                               }else {  taxaBrowser.addSeqToTree(name, taxon);  } //add it once
+                               }
+                               in.close();
                        }
-                       in.close();
-       
-                       taxaBrowser.assignHeirarchyIDs(0);
-                       
-                       if (m->control_pressed) {  for (int i = 0; i < outputNames.size(); i++) {       remove(outputNames[i].c_str()); } remove(tempTaxonomyFile.c_str()); delete classify; return 0; }
-
-                       taxaBrowser.binUnclassified();
-                       
                        remove(tempTaxonomyFile.c_str());
                        
                        if (m->control_pressed) {  for (int i = 0; i < outputNames.size(); i++) {       remove(outputNames[i].c_str()); } delete classify; return 0; }
-
                        
                        //print summary file
                        ofstream outTaxTree;
                        openOutputFile(taxSummary, outTaxTree);
-                       taxaBrowser.print(outTaxTree);
+                       taxaSum.print(outTaxTree);
                        outTaxTree.close();
                        
                        //output taxonomy with the unclassified bins added
@@ -533,9 +661,10 @@ int ClassifySeqsCommand::execute(){
                        openOutputFile(unclass, outTax);
                        
                        //get maxLevel from phylotree so you know how many 'unclassified's to add
-                       int maxLevel = taxaBrowser.getMaxLevel();
+                       int maxLevel = taxaSum.getMaxLevel();
                        
                        //read taxfile - this reading and rewriting is done to preserve the confidence scores.
+                       string name, taxon;
                        while (!inTax.eof()) {
                                if (m->control_pressed) {  for (int i = 0; i < outputNames.size(); i++) {       remove(outputNames[i].c_str()); } remove(unclass.c_str()); delete classify; return 0; }
 
@@ -551,6 +680,9 @@ int ClassifySeqsCommand::execute(){
                        remove(newTaxonomyFile.c_str());
                        rename(unclass.c_str(), newTaxonomyFile.c_str());
                        
+                       m->mothurOutEndLine();
+                       m->mothurOut("It took " + toString(time(NULL) - start) + " secs to create the summary file for " + toString(numFastaSeqs) + " sequences."); m->mothurOutEndLine(); m->mothurOutEndLine();
+                       
                        #ifdef USE_MPI  
                                }
                        #endif
@@ -559,10 +691,6 @@ int ClassifySeqsCommand::execute(){
                        m->mothurOut("Output File Names: "); m->mothurOutEndLine();
                        for (int i = 0; i < outputNames.size(); i++) {  m->mothurOut(outputNames[i]); m->mothurOutEndLine();    }
                        m->mothurOutEndLine();
-
-                       
-                       m->mothurOutEndLine();
-                       m->mothurOut("It took " + toString(time(NULL) - start) + " secs to classify " + toString(numFastaSeqs) + " sequences."); m->mothurOutEndLine(); m->mothurOutEndLine();
                }
                
                delete classify;
@@ -679,9 +807,9 @@ int ClassifySeqsCommand::driver(linePair* line, string taxFName, string tempTFNa
 
                for(int i=0;i<line->numSeqs;i++){
                        if (m->control_pressed) { return 0; }
-                       
-                       Sequence* candidateSeq = new Sequence(inFASTA);
-                       
+               
+                       Sequence* candidateSeq = new Sequence(inFASTA); gobble(inFASTA);
+               
                        if (candidateSeq->getName() != "") {
                                taxonomy = classify->getTaxonomy(candidateSeq);
                                
@@ -718,7 +846,7 @@ int ClassifySeqsCommand::driver(linePair* line, string taxFName, string tempTFNa
 }
 //**********************************************************************************************************************
 #ifdef USE_MPI
-int ClassifySeqsCommand::driverMPI(int start, int num, MPI_File& inMPI, MPI_File& newFile, MPI_File& tempFile, vector<long>& MPIPos){
+int ClassifySeqsCommand::driverMPI(int start, int num, MPI_File& inMPI, MPI_File& newFile, MPI_File& tempFile, vector<unsigned long int>& MPIPos){
        try {
                MPI_Status statusNew; 
                MPI_Status statusTemp; 
@@ -758,7 +886,7 @@ int ClassifySeqsCommand::driverMPI(int start, int num, MPI_File& inMPI, MPI_File
                                        }
                                        
                                        int length = outputString.length();
-                                       char* buf2 = new char[length];\r
+                                       char* buf2 = new char[length];
                                        memcpy(buf2, outputString.c_str(), length);
                                
                                        MPI_File_write_shared(newFile, buf2, length, MPI_CHAR, &statusNew);
@@ -766,7 +894,7 @@ int ClassifySeqsCommand::driverMPI(int start, int num, MPI_File& inMPI, MPI_File
 
                                        outputString =  candidateSeq->getName() + "\t" + classify->getSimpleTax() + "\n";
                                        length = outputString.length();
-                                       char* buf = new char[length];\r
+                                       char* buf = new char[length];
                                        memcpy(buf, outputString.c_str(), length);
                                
                                        MPI_File_write_shared(tempFile, buf, length, MPI_CHAR, &statusTemp);
@@ -799,12 +927,15 @@ int ClassifySeqsCommand::MPIReadNamesFile(string nameFilename){
                MPI_Offset size;
                MPI_Status status;
 
-               char* inFileName = new char[nameFilename.length()];\r
-               memcpy(inFileName, nameFilename.c_str(), nameFilename.length());
+               //char* inFileName = new char[nameFilename.length()];
+               //memcpy(inFileName, nameFilename.c_str(), nameFilename.length());
+               
+               char inFileName[1024];
+               strcpy(inFileName, nameFilename.c_str());
 
                MPI_File_open(MPI_COMM_WORLD, inFileName, MPI_MODE_RDONLY, MPI_INFO_NULL, &inMPI);  
                MPI_File_get_size(inMPI, &size);
-               delete inFileName;
+               //delete inFileName;
 
                char* buffer = new char[size];
                MPI_File_read(inMPI, buffer, size, MPI_CHAR, &status);
@@ -817,7 +948,11 @@ int ClassifySeqsCommand::MPIReadNamesFile(string nameFilename){
                string firstCol, secondCol;
                while(!iss.eof()) {
                        iss >> firstCol >> secondCol; gobble(iss);
-                       nameMap[firstCol] = getNumNames(secondCol);  //ex. seq1 seq1,seq3,seq5 -> seq1 = 3.
+                       
+                       vector<string> temp;
+                       splitAtComma(secondCol, temp);
+                       
+                       nameMap[firstCol] = temp;  
                }
        
                MPI_File_close(&inMPI);