1 #ifndef CLASSIFYSEQSCOMMAND_H
2 #define CLASSIFYSEQSCOMMAND_H
5 * classifyseqscommand.h
8 * Created by westcott on 11/2/09.
9 * Copyright 2009 Schloss Lab. All rights reserved.
14 #include "command.hpp"
16 #include "referencedb.h"
17 #include "sequence.hpp"
19 #include "phylotree.h"
20 #include "phylosummary.h"
24 //KNN and Bayesian methods modeled from algorithms in
25 //Naı¨ve Bayesian Classifier for Rapid Assignment of rRNA Sequences
26 //into the New Bacterial Taxonomy†
27 //Qiong Wang,1 George M. Garrity,1,2 James M. Tiedje,1,2 and James R. Cole1*
28 //Center for Microbial Ecology1 and Department of Microbiology and Molecular Genetics,2 Michigan State University,
29 //East Lansing, Michigan 48824
30 //Received 10 January 2007/Accepted 18 June 2007
34 class ClassifySeqsCommand : public Command {
37 ClassifySeqsCommand(string);
38 ClassifySeqsCommand();
39 ~ClassifySeqsCommand();
41 vector<string> setParameters();
42 string getCommandName() { return "classify.seqs"; }
43 string getCommandCategory() { return "Phylotype Analysis"; }
44 string getHelpString();
45 string getCitation() { return "Wang Q, Garrity GM, Tiedje JM, Cole JR (2007). Naive Bayesian classifier for rapid assignment of rRNA sequences into the new bacterial taxonomy. Appl Environ Microbiol 73: 5261-7. [ for Bayesian classifier ] \nAltschul SF, Madden TL, Schaffer AA, Zhang J, Zhang Z, Miller W, Lipman DJ (1997). Gapped BLAST and PSI-BLAST: a new generation of protein database search programs. Nucleic Acids Res 25: 3389-402. [ for BLAST ] \nDeSantis TZ, Hugenholtz P, Larsen N, Rojas M, Brodie EL, Keller K, Huber T, Dalevi D, Hu P, Andersen GL (2006). Greengenes, a chimera-checked 16S rRNA gene database and workbench compatible with ARB. Appl Environ Microbiol 72: 5069-72. [ for kmer ] \nhttp://www.mothur.org/wiki/Classify.seqs"; }
46 string getDescription() { return "classify sequences"; }
49 void help() { m->mothurOut(getHelpString()); }
55 unsigned long long start;
56 unsigned long long end;
57 linePair(unsigned long long i, unsigned long long j) : start(i), end(j) {}
60 vector<int> processIDS; //processid
61 vector<linePair*> lines;
62 vector<string> fastaFileNames;
63 vector<string> namefileNames;
64 vector<string> groupfileNames;
65 vector<string> outputNames;
66 map<string, vector<string> > nameMap;
67 map<string, vector<string> >::iterator itNames;
72 string fastaFileName, templateFileName, distanceFileName, namefile, search, method, taxonomyFileName, outputDir, groupfile;
73 int processors, kmerSize, numWanted, cutoff, iters;
74 float match, misMatch, gapOpen, gapExtend;
75 bool abort, probs, save, flip;
77 int driver(linePair*, string, string, string, string);
78 void appendTaxFiles(string, string);
79 int createProcesses(string, string, string, string);
80 string addUnclassifieds(string, int);
82 int MPIReadNamesFile(string);
84 int driverMPI(int, int, MPI_File&, MPI_File&, MPI_File&, MPI_File&, vector<unsigned long long>&);
88 /**************************************************************************************************/
89 //custom data structure for threads to use.
90 // This is passed by void pointer so it can be any data type
91 // that can be passed using a single void pointer (LPVOID).
96 string search, taxonomyFileName, templateFileName, method, accnos;
97 unsigned long long start;
98 unsigned long long end;
100 float match, misMatch, gapOpen, gapExtend;
101 int count, kmerSize, threadID, cutoff, iters, numWanted;
105 classifyData(string acc, bool p, string me, string te, string tx, string a, string r, string f, string se, int ks, int i, int numW, MothurOut* mout, unsigned long long st, unsigned long long en, float ma, float misMa, float gapO, float gapE, int cut, int tid, bool fli) {
107 taxonomyFileName = tx;
108 templateFileName = te;
132 /**************************************************************************************************/
133 #if defined (__APPLE__) || (__MACH__) || (linux) || (__linux)
135 static DWORD WINAPI MyClassThreadFunction(LPVOID lpParam){
136 classifyData* pDataArray;
137 pDataArray = (classifyData*)lpParam;
141 pDataArray->m->openOutputFile(pDataArray->taxFName, outTax);
143 ofstream outTaxSimple;
144 pDataArray->m->openOutputFile(pDataArray->tempTFName, outTaxSimple);
147 pDataArray->m->openOutputFile(pDataArray->accnos, outAcc);
150 pDataArray->m->openInputFile(pDataArray->filename, inFASTA);
154 //print header if you are process 0
155 if ((pDataArray->start == 0) || (pDataArray->start == 1)) {
157 }else { //this accounts for the difference in line endings.
158 inFASTA.seekg(pDataArray->start-1); pDataArray->m->gobble(inFASTA);
161 pDataArray->count = pDataArray->end;
164 Classify* myclassify;
165 if(pDataArray->method == "bayesian"){ myclassify = new Bayesian(pDataArray->taxonomyFileName, pDataArray->templateFileName, pDataArray->search, pDataArray->kmerSize, pDataArray->cutoff, pDataArray->iters, pDataArray->threadID, pDataArray->flip); }
166 else if(pDataArray->method == "knn"){ myclassify = new Knn(pDataArray->taxonomyFileName, pDataArray->templateFileName, pDataArray->search, pDataArray->kmerSize, pDataArray->gapOpen, pDataArray->gapExtend, pDataArray->match, pDataArray->misMatch, pDataArray->numWanted, pDataArray->threadID, pDataArray->flipThreshold); }
168 pDataArray->m->mothurOut(pDataArray->search + " is not a valid method option. I will run the command using bayesian.");
169 pDataArray->m->mothurOutEndLine();
170 myclassify = new Bayesian(pDataArray->taxonomyFileName, pDataArray->templateFileName, pDataArray->search, pDataArray->kmerSize, pDataArray->cutoff, pDataArray->iters, pDataArray->threadID, pDataArray->flip);
173 if (pDataArray->m->control_pressed) { delete myclassify; return 0; }
176 for(int i = 0; i < pDataArray->end; i++){ //end is the number of sequences to process
178 if (pDataArray->m->control_pressed) { delete myclassify; return 0; }
180 Sequence* candidateSeq = new Sequence(inFASTA); pDataArray->m->gobble(inFASTA);
182 if (candidateSeq->getName() != "") {
184 taxonomy = myclassify->getTaxonomy(candidateSeq);
186 if (pDataArray->m->control_pressed) { delete candidateSeq; return 0; }
188 if (taxonomy == "unknown;") { pDataArray->m->mothurOut("[WARNING]: " + candidateSeq->getName() + " could not be classified. You can use the remove.lineage command with taxon=unknown; to remove such sequences."); pDataArray->m->mothurOutEndLine(); }
190 //output confidence scores or not
191 if (pDataArray->probs) {
192 outTax << candidateSeq->getName() << '\t' << taxonomy << endl;
194 outTax << candidateSeq->getName() << '\t' << myclassify->getSimpleTax() << endl;
197 outTaxSimple << candidateSeq->getName() << '\t' << myclassify->getSimpleTax() << endl;
199 if (myclassify->getFlipped()) { outAcc << candidateSeq->getName() << endl; }
205 if((count) % 100 == 0){ pDataArray->m->mothurOut("Processing sequence: " + toString(count)); pDataArray->m->mothurOutEndLine(); }
209 if((count) % 100 != 0){ pDataArray->m->mothurOut("Processing sequence: " + toString(count)); pDataArray->m->mothurOutEndLine(); }
214 outTaxSimple.close();
217 catch(exception& e) {
218 pDataArray->m->errorOut(e, "ClassifySeqsCommand", "MyClassThreadFunction");