X-Git-Url: https://git.donarmstrong.com/?p=mothur.git;a=blobdiff_plain;f=classifyotucommand.cpp;h=76d7504aa04b0170629b0f022a882be94ceacfc0;hp=ec22de8b35762adf92e2a3b750198dcbb91f0396;hb=615301e57c25e241356a9c2380648d117709458d;hpb=bdd77df6a7df8a5775b57d0a7ec8b55de45e7b5b diff --git a/classifyotucommand.cpp b/classifyotucommand.cpp index ec22de8..76d7504 100644 --- a/classifyotucommand.cpp +++ b/classifyotucommand.cpp @@ -10,21 +10,24 @@ #include "classifyotucommand.h" #include "phylotree.h" #include "phylosummary.h" +#include "sharedutilities.h" //********************************************************************************************************************** vector ClassifyOtuCommand::setParameters(){ try { - CommandParameter plist("list", "InputTypes", "", "", "none", "none", "none",false,true); parameters.push_back(plist); - CommandParameter ptaxonomy("taxonomy", "InputTypes", "", "", "none", "none", "none",false,true); parameters.push_back(ptaxonomy); - CommandParameter preftaxonomy("reftaxonomy", "InputTypes", "", "", "none", "none", "none",false,false); parameters.push_back(preftaxonomy); - CommandParameter pname("name", "InputTypes", "", "", "none", "none", "none",false,false); parameters.push_back(pname); - CommandParameter pgroup("group", "InputTypes", "", "", "none", "none", "none",false,false); parameters.push_back(pgroup); - CommandParameter plabel("label", "String", "", "", "", "", "",false,false); parameters.push_back(plabel); - CommandParameter pbasis("basis", "Multiple", "otu-sequence", "otu", "", "", "",false,false); parameters.push_back(pbasis); - CommandParameter pcutoff("cutoff", "Number", "", "51", "", "", "",false,true); parameters.push_back(pcutoff); - CommandParameter pprobs("probs", "Boolean", "", "T", "", "", "",false,false); parameters.push_back(pprobs); - CommandParameter pinputdir("inputdir", "String", "", "", "", "", "",false,false); parameters.push_back(pinputdir); - CommandParameter poutputdir("outputdir", "String", "", "", "", "", "",false,false); parameters.push_back(poutputdir); + CommandParameter plist("list", "InputTypes", "", "", "none", "none", "none","",false,true,true); parameters.push_back(plist); + CommandParameter ptaxonomy("taxonomy", "InputTypes", "", "", "none", "none", "none","constaxonomy",false,true,true); parameters.push_back(ptaxonomy); + CommandParameter preftaxonomy("reftaxonomy", "InputTypes", "", "", "none", "none", "none","",false,false); parameters.push_back(preftaxonomy); + CommandParameter pname("name", "InputTypes", "", "", "NameCount", "none", "none","",false,false,true); parameters.push_back(pname); + CommandParameter pcount("count", "InputTypes", "", "", "NameCount-CountGroup", "none", "none","",false,false,true); parameters.push_back(pcount); + CommandParameter pgroup("group", "InputTypes", "", "", "CountGroup", "none", "none","",false,false,true); parameters.push_back(pgroup); + CommandParameter ppersample("persample", "Boolean", "", "F", "", "", "","",false,false); parameters.push_back(ppersample); + CommandParameter plabel("label", "String", "", "", "", "", "","",false,false); parameters.push_back(plabel); + CommandParameter pbasis("basis", "Multiple", "otu-sequence", "otu", "", "", "","",false,false); parameters.push_back(pbasis); + CommandParameter pcutoff("cutoff", "Number", "", "51", "", "", "","",false,true); parameters.push_back(pcutoff); + CommandParameter pprobs("probs", "Boolean", "", "T", "", "", "","",false,false); parameters.push_back(pprobs); + CommandParameter pinputdir("inputdir", "String", "", "", "", "", "","",false,false); parameters.push_back(pinputdir); + CommandParameter poutputdir("outputdir", "String", "", "", "", "", "","",false,false); parameters.push_back(poutputdir); vector myArray; for (int i = 0; i < parameters.size(); i++) { myArray.push_back(parameters[i].name); } @@ -39,16 +42,18 @@ vector ClassifyOtuCommand::setParameters(){ string ClassifyOtuCommand::getHelpString(){ try { string helpString = ""; - helpString += "The classify.otu command parameters are list, taxonomy, reftaxonomy, name, group, cutoff, label, basis and probs. The taxonomy and list parameters are required unless you have a valid current file.\n"; + helpString += "The classify.otu command parameters are list, taxonomy, reftaxonomy, name, group, count, persample, cutoff, label, basis and probs. The taxonomy and list parameters are required unless you have a valid current file.\n"; helpString += "The reftaxonomy parameter allows you give the name of the reference taxonomy file used when you classified your sequences. Providing it will keep the rankIDs in the summary file static.\n"; helpString += "The name parameter allows you add a names file with your taxonomy file.\n"; helpString += "The group parameter allows you provide a group file to use in creating the summary file breakdown.\n"; - helpString += "The basis parameter allows you indicate what you want the summary file to represent, options are otu and sequence. Default is otu.\n"; + helpString += "The count parameter allows you add a count file associated with your list file. When using the count parameter mothur assumes your list file contains only uniques.\n"; + helpString += "The basis parameter allows you indicate what you want the summary file to represent, options are otu and sequence. Default is otu.\n"; helpString += "For example consider the following basis=sequence could give Clostridiales 3 105 16 43 46, where 105 is the total number of sequences whose otu classified to Clostridiales.\n"; helpString += "16 is the number of sequences in the otus from groupA, 43 is the number of sequences in the otus from groupB, and 46 is the number of sequences in the otus from groupC.\n"; helpString += "Now for basis=otu could give Clostridiales 3 7 6 1 2, where 7 is the number of otus that classified to Clostridiales.\n"; helpString += "6 is the number of otus containing sequences from groupA, 1 is the number of otus containing sequences from groupB, and 2 is the number of otus containing sequences from groupC.\n"; helpString += "The label parameter allows you to select what distance levels you would like a output files created for, and is separated by dashes.\n"; + helpString += "The persample parameter allows you to find a consensus taxonomy for each group. Default=f\n"; helpString += "The default value for label is all labels in your inputfile.\n"; helpString += "The cutoff parameter allows you to specify a consensus confidence threshold for your taxonomy. The default is 51, meaning 51%. Cutoff cannot be below 51.\n"; helpString += "The probs parameter shuts off the outputting of the consensus confidence results. The default is true, meaning you want the confidence to be shown.\n"; @@ -63,6 +68,22 @@ string ClassifyOtuCommand::getHelpString(){ } } //********************************************************************************************************************** +string ClassifyOtuCommand::getOutputPattern(string type) { + try { + string pattern = ""; + + if (type == "constaxonomy") { pattern = "[filename],[distance],cons.taxonomy"; } + else if (type == "taxsummary") { pattern = "[filename],[distance],cons.tax.summary"; } + else { m->mothurOut("[ERROR]: No definition for type " + type + " output pattern.\n"); m->control_pressed = true; } + + return pattern; + } + catch(exception& e) { + m->errorOut(e, "ClassifyOtuCommand", "getOutputPattern"); + exit(1); + } +} +//********************************************************************************************************************** ClassifyOtuCommand::ClassifyOtuCommand(){ try { abort = true; calledHelp = true; @@ -151,6 +172,14 @@ ClassifyOtuCommand::ClassifyOtuCommand(string option) { //if the user has not given a path then, add inputdir. else leave path alone. if (path == "") { parameters["group"] = inputDir + it->second; } } + + it = parameters.find("count"); + //user has given a template file + if(it != parameters.end()){ + path = m->hasPath(it->second); + //if the user has not given a path then, add inputdir. else leave path alone. + if (path == "") { parameters["count"] = inputDir + it->second; } + } } @@ -190,6 +219,20 @@ ClassifyOtuCommand::ClassifyOtuCommand(string option) { if (groupfile == "not open") { abort = true; } else if (groupfile == "not found") { groupfile = ""; } else { m->setGroupFile(groupfile); } + + countfile = validParameter.validFile(parameters, "count", true); + if (countfile == "not open") { countfile = ""; abort = true; } + else if (countfile == "not found") { countfile = ""; } + else { m->setCountTableFile(countfile); } + + if ((namefile != "") && (countfile != "")) { + m->mothurOut("[ERROR]: you may only use one of the following: name or count."); m->mothurOutEndLine(); abort = true; + } + + if ((groupfile != "") && (countfile != "")) { + m->mothurOut("[ERROR]: you may only use one of the following: group or count."); m->mothurOutEndLine(); abort=true; + } + //check for optional parameter and set defaults // ...at some point should added some additional type checking... @@ -199,7 +242,7 @@ ClassifyOtuCommand::ClassifyOtuCommand(string option) { if(label != "all") { m->splitAtDash(label, labels); allLines = 0; } else { allLines = 1; } } - + basis = validParameter.validFile(parameters, "basis", false); if (basis == "not found") { basis = "otu"; } @@ -210,14 +253,27 @@ ClassifyOtuCommand::ClassifyOtuCommand(string option) { temp = validParameter.validFile(parameters, "probs", false); if (temp == "not found"){ temp = "true"; } probs = m->isTrue(temp); - + + temp = validParameter.validFile(parameters, "persample", false); if (temp == "not found"){ temp = "f"; } + persample = m->isTrue(temp); + + if ((groupfile == "") && (countfile == "")) { if (persample) { m->mothurOut("persample is only valid with a group file, or count file with group information. Setting persample=f.\n"); persample = false; } + } + if (countfile != "") { + CountTable cts; + if (!cts.testGroups(countfile)) { + if (persample) { m->mothurOut("persample is only valid with a group file, or count file with group information. Setting persample=f.\n"); persample = false; } + } + } if ((cutoff < 51) || (cutoff > 100)) { m->mothurOut("cutoff must be above 50, and no greater than 100."); m->mothurOutEndLine(); abort = true; } - if (namefile == ""){ - vector files; files.push_back(taxfile); - parser.getNameFile(files); - } + if (countfile == "") { + if (namefile == ""){ + vector files; files.push_back(taxfile); + parser.getNameFile(files); + } + } } } @@ -234,10 +290,14 @@ int ClassifyOtuCommand::execute(){ if (abort == true) { if (calledHelp) { return 0; } return 2; } //if user gave a namesfile then use it - if (namefile != "") { readNamesFile(); } - + if (namefile != "") { m->readNames(namefile, nameMap, true); } + if (groupfile != "") { groupMap = new GroupMap(groupfile); groupMap->readMap(); groups = groupMap->getNamesOfGroups(); } + else { groupMap = NULL; } + if (countfile != "") { ct = new CountTable(); ct->readTable(countfile, true); if (ct->hasGroupInfo()) { groups = ct->getNamesOfGroups(); } } + else { ct = NULL; } + //read taxonomy file and save in map for easy access in building bin trees - readTaxonomyFile(); + m->readTax(taxfile, taxMap); if (m->control_pressed) { return 0; } @@ -249,7 +309,7 @@ int ClassifyOtuCommand::execute(){ set processedLabels; set userLabels = labels; - if (m->control_pressed) { outputTypes.clear(); delete input; delete list; for (int i = 0; i < outputNames.size(); i++) { m->mothurRemove(outputNames[i]); } return 0; } + if (m->control_pressed) { outputTypes.clear(); if (ct != NULL) { delete ct; } if (groupMap != NULL) { delete groupMap; } delete input; delete list; for (int i = 0; i < outputNames.size(); i++) { m->mothurRemove(outputNames[i]); } return 0; } while((list != NULL) && ((allLines == 1) || (userLabels.size() != 0))) { @@ -257,7 +317,7 @@ int ClassifyOtuCommand::execute(){ m->mothurOut(list->getLabel() + "\t" + toString(list->size())); m->mothurOutEndLine(); process(list); - if (m->control_pressed) { outputTypes.clear(); for (int i = 0; i < outputNames.size(); i++) { m->mothurRemove(outputNames[i]); } delete input; delete list; return 0; } + if (m->control_pressed) { outputTypes.clear(); for (int i = 0; i < outputNames.size(); i++) { m->mothurRemove(outputNames[i]); } if (ct != NULL) { delete ct; } if (groupMap != NULL) { delete groupMap; } delete input; delete list; return 0; } processedLabels.insert(list->getLabel()); userLabels.erase(list->getLabel()); @@ -272,7 +332,7 @@ int ClassifyOtuCommand::execute(){ process(list); - if (m->control_pressed) { outputTypes.clear(); for (int i = 0; i < outputNames.size(); i++) { m->mothurRemove(outputNames[i]); } delete input; delete list; return 0; } + if (m->control_pressed) { outputTypes.clear(); if (ct != NULL) { delete ct; } if (groupMap != NULL) { delete groupMap; } for (int i = 0; i < outputNames.size(); i++) { m->mothurRemove(outputNames[i]); } delete input; delete list; return 0; } processedLabels.insert(list->getLabel()); userLabels.erase(list->getLabel()); @@ -308,10 +368,12 @@ int ClassifyOtuCommand::execute(){ process(list); delete list; - if (m->control_pressed) { outputTypes.clear(); for (int i = 0; i < outputNames.size(); i++) { m->mothurRemove(outputNames[i]); } delete input; delete list; return 0; } + if (m->control_pressed) { outputTypes.clear(); if (ct != NULL) { delete ct; } if (groupMap != NULL) { delete groupMap; } for (int i = 0; i < outputNames.size(); i++) { m->mothurRemove(outputNames[i]); } delete input; delete list; return 0; } } delete input; + if (groupMap != NULL) { delete groupMap; } + if (ct != NULL) { delete ct; } if (m->control_pressed) { outputTypes.clear(); for (int i = 0; i < outputNames.size(); i++) { m->mothurRemove(outputNames[i]); } return 0; } @@ -327,80 +389,14 @@ int ClassifyOtuCommand::execute(){ exit(1); } } - //********************************************************************************************************************** -int ClassifyOtuCommand::readNamesFile() { - try { - - ifstream inNames; - m->openInputFile(namefile, inNames); - - string name, names; - - while(!inNames.eof()){ - inNames >> name; //read from first column A - inNames >> names; //read from second column A,B,C,D - m->gobble(inNames); - - //parse names into vector - vector theseNames; - m->splitAtComma(names, theseNames); - - for (int i = 0; i < theseNames.size(); i++) { nameMap[theseNames[i]] = name; } - - if (m->control_pressed) { inNames.close(); nameMap.clear(); return 0; } - } - inNames.close(); - - return 0; - } - catch(exception& e) { - m->errorOut(e, "ClassifyOtuCommand", "readNamesFile"); - exit(1); - } -} -//********************************************************************************************************************** -int ClassifyOtuCommand::readTaxonomyFile() { - try { - - ifstream in; - m->openInputFile(taxfile, in); - - string name, tax; - - while(!in.eof()){ - in >> name >> tax; - m->gobble(in); - - //are there confidence scores, if so remove them - if (tax.find_first_of('(') != -1) { m->removeConfidences(tax); } - - taxMap[name] = tax; - - if (m->control_pressed) { in.close(); taxMap.clear(); return 0; } - } - in.close(); - - return 0; - } - catch(exception& e) { - m->errorOut(e, "ClassifyOtuCommand", "readTaxonomyFile"); - exit(1); - } -} -//********************************************************************************************************************** -vector ClassifyOtuCommand::findConsensusTaxonomy(int bin, ListVector* thisList, int& size, string& conTax) { +vector ClassifyOtuCommand::findConsensusTaxonomy(vector names, int& size, string& conTax) { try{ conTax = ""; - vector names; vector allNames; map::iterator it; map::iterator it2; - //parse names into vector - string binnames = thisList->get(bin); - m->splitAtComma(binnames, names); - //create a tree containing sequences from this bin PhyloTree* phylo = new PhyloTree(); @@ -440,10 +436,16 @@ vector ClassifyOtuCommand::findConsensusTaxonomy(int bin, ListVector* th if (it == taxMap.end()) { //this name is not in taxonomy file, skip it m->mothurOut(names[i] + " is not in your taxonomy file. I will not include it in the consensus."); m->mothurOutEndLine(); }else{ + if (countfile != "") { + int numDups = ct->getNumSeqs(names[i]); + for (int j = 0; j < numDups; j++) { phylo->addSeqToTree(names[i], it->second); } + size += numDups; + }else{ //add seq to tree - phylo->addSeqToTree(names[i], it->second); - size++; - allNames.push_back(names[i]); + phylo->addSeqToTree(names[i], it->second); + size++; + } + allNames.push_back(names[i]); } } @@ -526,46 +528,189 @@ int ClassifyOtuCommand::process(ListVector* processList) { if (outputDir == "") { outputDir += m->hasPath(listfile); } ofstream out; - string outputFile = outputDir + m->getRootName(m->getSimpleName(listfile)) + processList->getLabel() + ".cons.taxonomy"; + map variables; + variables["[filename]"] = outputDir + m->getRootName(m->getSimpleName(listfile)); + variables["[distance]"] = processList->getLabel(); + string outputFile = getOutputFileName("constaxonomy", variables); m->openOutputFile(outputFile, out); outputNames.push_back(outputFile); outputTypes["constaxonomy"].push_back(outputFile); ofstream outSum; - string outputSumFile = outputDir + m->getRootName(m->getSimpleName(listfile)) + processList->getLabel() + ".cons.tax.summary"; + string outputSumFile = getOutputFileName("taxsummary", variables); m->openOutputFile(outputSumFile, outSum); outputNames.push_back(outputSumFile); outputTypes["taxsummary"].push_back(outputSumFile); out << "OTU\tSize\tTaxonomy" << endl; PhyloSummary* taxaSum; - if (refTaxonomy != "") { - taxaSum = new PhyloSummary(refTaxonomy, groupfile); + if (countfile != "") { + if (refTaxonomy != "") { taxaSum = new PhyloSummary(refTaxonomy, ct); } + else { taxaSum = new PhyloSummary(ct); } }else { - taxaSum = new PhyloSummary(groupfile); - } - + if (refTaxonomy != "") { taxaSum = new PhyloSummary(refTaxonomy, groupMap); } + else { taxaSum = new PhyloSummary(groupMap); } + } + + vector outSums; + vector outs; + vector taxaSums; + map groupIndex; + if (persample) { + for (int i = 0; i < groups.size(); i++) { + groupIndex[groups[i]] = i; + ofstream* temp = new ofstream(); + variables["[distance]"] = processList->getLabel() + "." + groups[i]; + string outputFile = getOutputFileName("constaxonomy", variables); + m->openOutputFile(outputFile, *temp); + (*temp) << "OTU\tSize\tTaxonomy" << endl; + outs.push_back(temp); + outputNames.push_back(outputFile); outputTypes["constaxonomy"].push_back(outputFile); + + ofstream* tempSum = new ofstream(); + string outputSumFile = getOutputFileName("taxsummary", variables); + m->openOutputFile(outputSumFile, *tempSum); + outSums.push_back(tempSum); + outputNames.push_back(outputSumFile); outputTypes["taxsummary"].push_back(outputSumFile); + + PhyloSummary* taxaSumt; + if (countfile != "") { + if (refTaxonomy != "") { taxaSumt = new PhyloSummary(refTaxonomy, ct); } + else { taxaSumt = new PhyloSummary(ct); } + }else { + if (refTaxonomy != "") { taxaSumt = new PhyloSummary(refTaxonomy, groupMap); } + else { taxaSumt = new PhyloSummary(groupMap); } + } + taxaSums.push_back(taxaSumt); + } + } + //for each bin in the list vector + string snumBins = toString(processList->getNumBins()); for (int i = 0; i < processList->getNumBins(); i++) { if (m->control_pressed) { break; } vector names; - names = findConsensusTaxonomy(i, processList, size, conTax); + string binnames = processList->get(i); + vector thisNames; + m->splitAtComma(binnames, thisNames); + + names = findConsensusTaxonomy(thisNames, size, conTax); - if (m->control_pressed) { out.close(); return 0; } + if (m->control_pressed) { break; } //output to new names file - out << (i+1) << '\t' << size << '\t' << conTax << endl; + string binLabel = "Otu"; + string sbinNumber = toString(i+1); + if (sbinNumber.length() < snumBins.length()) { + int diff = snumBins.length() - sbinNumber.length(); + for (int h = 0; h < diff; h++) { binLabel += "0"; } + } + binLabel += sbinNumber; + + out << binLabel << '\t' << size << '\t' << conTax << endl; string noConfidenceConTax = conTax; m->removeConfidences(noConfidenceConTax); //add this bins taxonomy to summary if (basis == "sequence") { - for(int j = 0; j < names.size(); j++) { taxaSum->addSeqToTree(names[j], noConfidenceConTax); } + for(int j = 0; j < names.size(); j++) { + //int numReps = 1; + //if (countfile != "") { numReps = ct->getNumSeqs(names[j]); } + //for(int k = 0; k < numReps; k++) { taxaSum->addSeqToTree(names[j], noConfidenceConTax); } + taxaSum->addSeqToTree(names[j], noConfidenceConTax); + } }else { //otu - taxaSum->addSeqToTree(noConfidenceConTax, names); + map containsGroup; + if (countfile != "") { + if (ct->hasGroupInfo()) { + vector mGroups = ct->getNamesOfGroups(); + for (int k = 0; k < names.size(); k++) { + vector counts = ct->getGroupCounts(names[k]); + for (int h = 0; h < counts.size(); h++) { + if (counts[h] != 0) { containsGroup[mGroups[h]] = true; } + } + } + } + }else { + if (groupfile != "") { + vector mGroups = groupMap->getNamesOfGroups(); + for (int j = 0; j < mGroups.size(); j++) { containsGroup[mGroups[j]] = false; } + + for (int k = 0; k < names.size(); k++) { + //find out the sequences group + string group = groupMap->getGroup(names[k]); + + if (group == "not found") { m->mothurOut("[WARNING]: " + names[k] + " is not in your groupfile, and will be included in the overall total, but not any group total."); m->mothurOutEndLine(); } + else { + containsGroup[group] = true; + } + } + } + } + taxaSum->addSeqToTree(noConfidenceConTax, containsGroup); } + + + if (persample) { + //divide names by group + map > parsedNames; + map >::iterator itParsed; + + //parse names by group + for (int j = 0; j < names.size(); j++) { + if (groupfile != "") { + string group = groupMap->getGroup(names[j]); + itParsed = parsedNames.find(group); + + if (itParsed != parsedNames.end()) { itParsed->second.push_back(names[j]); } + else { vector tempNames; tempNames.push_back(names[j]); parsedNames[group] = tempNames; } + }else { //count file was used + vector thisSeqsGroups = ct->getGroups(names[j]); + for (int k = 0; k < thisSeqsGroups.size(); k++) { + string group = thisSeqsGroups[k]; + itParsed = parsedNames.find(group); + + if (itParsed != parsedNames.end()) { itParsed->second.push_back(names[j]); } + else { vector tempNames; tempNames.push_back(names[j]); parsedNames[group] = tempNames; } + } + } + } + + for (itParsed = parsedNames.begin(); itParsed != parsedNames.end(); itParsed++) { + vector theseNames = findConsensusTaxonomy(itParsed->second, size, conTax); + + if (m->control_pressed) { break; } + + //output to new names file + string binLabel = "Otu"; + string sbinNumber = toString(i+1); + if (sbinNumber.length() < snumBins.length()) { + int diff = snumBins.length() - sbinNumber.length(); + for (int h = 0; h < diff; h++) { binLabel += "0"; } + } + binLabel += sbinNumber; + + (*outs[groupIndex[itParsed->first]]) << binLabel << '\t' << size << '\t' << conTax << endl; + + string noConfidenceConTax = conTax; + m->removeConfidences(noConfidenceConTax); + + //add this bins taxonomy to summary + if (basis == "sequence") { + for(int j = 0; j < theseNames.size(); j++) { + int numReps = 1; + if (countfile != "") { numReps = ct->getGroupCount(theseNames[j], itParsed->first); } //get num seqs for this seq from this group + for(int k = 0; k < numReps; k++) { (taxaSums[groupIndex[itParsed->first]])->addSeqToTree(theseNames[j], noConfidenceConTax); } + } + }else { //otu + map containsGroup; + containsGroup[itParsed->first] = true; + (taxaSums[groupIndex[itParsed->first]])->addSeqToTree(noConfidenceConTax, containsGroup); + } + } + } } out.close(); @@ -573,6 +718,17 @@ int ClassifyOtuCommand::process(ListVector* processList) { //print summary file taxaSum->print(outSum); outSum.close(); + + if (persample) { + for (int i = 0; i < groups.size(); i++) { + (*outs[i]).close(); + taxaSums[i]->print(*outSums[i]); + (*outSums[i]).close(); + delete outs[i]; + delete outSums[i]; + delete taxaSums[i]; + } + } delete taxaSum;