2 * classifyotucommand.cpp
5 * Created by westcott on 6/1/10.
6 * Copyright 2010 Schloss Lab. All rights reserved.
10 #include "classifyotucommand.h"
11 #include "phylotree.h"
12 #include "phylosummary.h"
14 //**********************************************************************************************************************
15 vector<string> ClassifyOtuCommand::setParameters(){
17 CommandParameter plist("list", "InputTypes", "", "", "none", "none", "none",false,true); parameters.push_back(plist);
18 CommandParameter ptaxonomy("taxonomy", "InputTypes", "", "", "none", "none", "none",false,true); parameters.push_back(ptaxonomy);
19 CommandParameter preftaxonomy("reftaxonomy", "InputTypes", "", "", "none", "none", "none",false,false); parameters.push_back(preftaxonomy);
20 CommandParameter pname("name", "InputTypes", "", "", "none", "none", "none",false,false); parameters.push_back(pname);
21 CommandParameter pgroup("group", "InputTypes", "", "", "none", "none", "none",false,false); parameters.push_back(pgroup);
22 CommandParameter plabel("label", "String", "", "", "", "", "",false,false); parameters.push_back(plabel);
23 CommandParameter pbasis("basis", "Multiple", "otu-sequence", "otu", "", "", "",false,false); parameters.push_back(pbasis);
24 CommandParameter pcutoff("cutoff", "Number", "", "51", "", "", "",false,true); parameters.push_back(pcutoff);
25 CommandParameter pprobs("probs", "Boolean", "", "T", "", "", "",false,false); parameters.push_back(pprobs);
26 CommandParameter pinputdir("inputdir", "String", "", "", "", "", "",false,false); parameters.push_back(pinputdir);
27 CommandParameter poutputdir("outputdir", "String", "", "", "", "", "",false,false); parameters.push_back(poutputdir);
29 vector<string> myArray;
30 for (int i = 0; i < parameters.size(); i++) { myArray.push_back(parameters[i].name); }
34 m->errorOut(e, "ClassifyOtuCommand", "setParameters");
38 //**********************************************************************************************************************
39 string ClassifyOtuCommand::getHelpString(){
41 string helpString = "";
42 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";
43 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";
44 helpString += "The name parameter allows you add a names file with your taxonomy file.\n";
45 helpString += "The group parameter allows you provide a group file to use in creating the summary file breakdown.\n";
46 helpString += "The basis parameter allows you indicate what you want the summary file to represent, options are otu and sequence. Default is otu.\n";
47 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";
48 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";
49 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";
50 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";
51 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";
52 helpString += "The default value for label is all labels in your inputfile.\n";
53 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";
54 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";
55 helpString += "The classify.otu command should be in the following format: classify.otu(taxonomy=yourTaxonomyFile, list=yourListFile, name=yourNamesFile, label=yourLabels).\n";
56 helpString += "Example classify.otu(taxonomy=abrecovery.silva.full.taxonomy, list=abrecovery.fn.list, label=0.10).\n";
57 helpString += "Note: No spaces between parameter labels (i.e. list), '=' and parameters (i.e.yourListFile).\n";
61 m->errorOut(e, "ClassifyOtuCommand", "getHelpString");
65 //**********************************************************************************************************************
66 ClassifyOtuCommand::ClassifyOtuCommand(){
68 abort = true; calledHelp = true;
70 vector<string> tempOutNames;
71 outputTypes["constaxonomy"] = tempOutNames;
72 outputTypes["taxsummary"] = tempOutNames;
75 m->errorOut(e, "ClassifyOtuCommand", "ClassifyOtuCommand");
80 //**********************************************************************************************************************
81 ClassifyOtuCommand::ClassifyOtuCommand(string option) {
83 abort = false; calledHelp = false;
87 //allow user to run help
88 if (option == "help") {
89 help(); abort = true; calledHelp = true;
90 }else if(option == "citation") { citation(); abort = true; calledHelp = true;}
92 vector<string> myArray = setParameters();
94 OptionParser parser(option);
95 map<string, string> parameters = parser.getParameters();
97 ValidParameters validParameter;
98 map<string, string>::iterator it;
100 //check to make sure all parameters are valid for command
101 for (it = parameters.begin(); it != parameters.end(); it++) {
102 if (validParameter.isValidParameter(it->first, myArray, it->second) != true) { abort = true; }
105 //initialize outputTypes
106 vector<string> tempOutNames;
107 outputTypes["constaxonomy"] = tempOutNames;
108 outputTypes["taxsummary"] = tempOutNames;
110 //if the user changes the input directory command factory will send this info to us in the output parameter
111 string inputDir = validParameter.validFile(parameters, "inputdir", false);
112 if (inputDir == "not found"){ inputDir = ""; }
115 it = parameters.find("list");
116 //user has given a template file
117 if(it != parameters.end()){
118 path = m->hasPath(it->second);
119 //if the user has not given a path then, add inputdir. else leave path alone.
120 if (path == "") { parameters["list"] = inputDir + it->second; }
123 it = parameters.find("name");
124 //user has given a template file
125 if(it != parameters.end()){
126 path = m->hasPath(it->second);
127 //if the user has not given a path then, add inputdir. else leave path alone.
128 if (path == "") { parameters["name"] = inputDir + it->second; }
131 it = parameters.find("taxonomy");
132 //user has given a template file
133 if(it != parameters.end()){
134 path = m->hasPath(it->second);
135 //if the user has not given a path then, add inputdir. else leave path alone.
136 if (path == "") { parameters["taxonomy"] = inputDir + it->second; }
139 it = parameters.find("reftaxonomy");
140 //user has given a template file
141 if(it != parameters.end()){
142 path = m->hasPath(it->second);
143 //if the user has not given a path then, add inputdir. else leave path alone.
144 if (path == "") { parameters["reftaxonomy"] = inputDir + it->second; }
147 it = parameters.find("group");
148 //user has given a template file
149 if(it != parameters.end()){
150 path = m->hasPath(it->second);
151 //if the user has not given a path then, add inputdir. else leave path alone.
152 if (path == "") { parameters["group"] = inputDir + it->second; }
157 //if the user changes the output directory command factory will send this info to us in the output parameter
158 outputDir = validParameter.validFile(parameters, "outputdir", false); if (outputDir == "not found"){ outputDir = ""; }
160 //check for required parameters
161 listfile = validParameter.validFile(parameters, "list", true);
162 if (listfile == "not found") {
163 //if there is a current list file, use it
164 listfile = m->getListFile();
165 if (listfile != "") { m->mothurOut("Using " + listfile + " as input file for the list parameter."); m->mothurOutEndLine(); }
166 else { m->mothurOut("You have no current listfile and the list parameter is required."); m->mothurOutEndLine(); abort = true; }
168 else if (listfile == "not open") { abort = true; }
169 else { m->setListFile(listfile); }
171 taxfile = validParameter.validFile(parameters, "taxonomy", true);
172 if (taxfile == "not found") { //if there is a current list file, use it
173 taxfile = m->getTaxonomyFile();
174 if (taxfile != "") { m->mothurOut("Using " + taxfile + " as input file for the taxonomy parameter."); m->mothurOutEndLine(); }
175 else { m->mothurOut("You have no current taxonomy file and the taxonomy parameter is required."); m->mothurOutEndLine(); abort = true; }
177 else if (taxfile == "not open") { abort = true; }
178 else { m->setTaxonomyFile(taxfile); }
180 refTaxonomy = validParameter.validFile(parameters, "reftaxonomy", true);
181 if (refTaxonomy == "not found") { refTaxonomy = ""; m->mothurOut("reftaxonomy is not required, but if given will keep the rankIDs in the summary file static."); m->mothurOutEndLine(); }
182 else if (refTaxonomy == "not open") { abort = true; }
184 namefile = validParameter.validFile(parameters, "name", true);
185 if (namefile == "not open") { namefile = ""; abort = true; }
186 else if (namefile == "not found") { namefile = ""; }
187 else { m->setNameFile(namefile); }
189 groupfile = validParameter.validFile(parameters, "group", true);
190 if (groupfile == "not open") { abort = true; }
191 else if (groupfile == "not found") { groupfile = ""; }
192 else { m->setGroupFile(groupfile); }
194 //check for optional parameter and set defaults
195 // ...at some point should added some additional type checking...
196 label = validParameter.validFile(parameters, "label", false);
197 if (label == "not found") { label = ""; allLines = 1; }
199 if(label != "all") { m->splitAtDash(label, labels); allLines = 0; }
200 else { allLines = 1; }
203 basis = validParameter.validFile(parameters, "basis", false);
204 if (basis == "not found") { basis = "otu"; }
206 if ((basis != "otu") && (basis != "sequence")) { m->mothurOut("Invalid option for basis. basis options are otu and sequence, using otu."); m->mothurOutEndLine(); }
208 string temp = validParameter.validFile(parameters, "cutoff", false); if (temp == "not found") { temp = "51"; }
209 m->mothurConvert(temp, cutoff);
211 temp = validParameter.validFile(parameters, "probs", false); if (temp == "not found"){ temp = "true"; }
212 probs = m->isTrue(temp);
215 if ((cutoff < 51) || (cutoff > 100)) { m->mothurOut("cutoff must be above 50, and no greater than 100."); m->mothurOutEndLine(); abort = true; }
218 vector<string> files; files.push_back(taxfile);
219 parser.getNameFile(files);
224 catch(exception& e) {
225 m->errorOut(e, "ClassifyOtuCommand", "ClassifyOtuCommand");
229 //**********************************************************************************************************************
231 int ClassifyOtuCommand::execute(){
234 if (abort == true) { if (calledHelp) { return 0; } return 2; }
236 //if user gave a namesfile then use it
237 if (namefile != "") { readNamesFile(); }
239 //read taxonomy file and save in map for easy access in building bin trees
242 if (m->control_pressed) { return 0; }
244 input = new InputData(listfile, "list");
245 list = input->getListVector();
246 string lastLabel = list->getLabel();
248 //if the users enters label "0.06" and there is no "0.06" in their file use the next lowest label.
249 set<string> processedLabels;
250 set<string> userLabels = labels;
252 if (m->control_pressed) { outputTypes.clear(); delete input; delete list; for (int i = 0; i < outputNames.size(); i++) { m->mothurRemove(outputNames[i]); } return 0; }
254 while((list != NULL) && ((allLines == 1) || (userLabels.size() != 0))) {
256 if (allLines == 1 || labels.count(list->getLabel()) == 1){
258 m->mothurOut(list->getLabel() + "\t" + toString(list->size())); m->mothurOutEndLine();
260 if (m->control_pressed) { outputTypes.clear(); for (int i = 0; i < outputNames.size(); i++) { m->mothurRemove(outputNames[i]); } delete input; delete list; return 0; }
262 processedLabels.insert(list->getLabel());
263 userLabels.erase(list->getLabel());
266 if ((m->anyLabelsToProcess(list->getLabel(), userLabels, "") == true) && (processedLabels.count(lastLabel) != 1)) {
267 string saveLabel = list->getLabel();
270 list = input->getListVector(lastLabel);
271 m->mothurOut(list->getLabel() + "\t" + toString(list->size())); m->mothurOutEndLine();
275 if (m->control_pressed) { outputTypes.clear(); for (int i = 0; i < outputNames.size(); i++) { m->mothurRemove(outputNames[i]); } delete input; delete list; return 0; }
277 processedLabels.insert(list->getLabel());
278 userLabels.erase(list->getLabel());
280 //restore real lastlabel to save below
281 list->setLabel(saveLabel);
284 lastLabel = list->getLabel();
287 list = input->getListVector();
290 //output error messages about any remaining user labels
291 bool needToRun = false;
292 for (set<string>::iterator it = userLabels.begin(); it != userLabels.end(); it++) {
293 m->mothurOut("Your file does not include the label " + (*it));
294 if (processedLabels.count(lastLabel) != 1) {
295 m->mothurOut(". I will use " + lastLabel + "."); m->mothurOutEndLine();
298 m->mothurOut(". Please refer to " + lastLabel + "."); m->mothurOutEndLine();
302 //run last label if you need to
303 if (needToRun == true) {
304 if (list != NULL) { delete list; }
305 list = input->getListVector(lastLabel);
306 m->mothurOut(list->getLabel() + "\t" + toString(list->size())); m->mothurOutEndLine();
311 if (m->control_pressed) { outputTypes.clear(); for (int i = 0; i < outputNames.size(); i++) { m->mothurRemove(outputNames[i]); } delete input; delete list; return 0; }
316 if (m->control_pressed) { outputTypes.clear(); for (int i = 0; i < outputNames.size(); i++) { m->mothurRemove(outputNames[i]); } return 0; }
318 m->mothurOutEndLine();
319 m->mothurOut("Output File Names: "); m->mothurOutEndLine();
320 for (int i = 0; i < outputNames.size(); i++) { m->mothurOut(outputNames[i]); m->mothurOutEndLine(); }
321 m->mothurOutEndLine();
325 catch(exception& e) {
326 m->errorOut(e, "ClassifyOtuCommand", "execute");
331 //**********************************************************************************************************************
332 int ClassifyOtuCommand::readNamesFile() {
336 m->openInputFile(namefile, inNames);
340 while(!inNames.eof()){
341 inNames >> name; //read from first column A
342 inNames >> names; //read from second column A,B,C,D
345 //parse names into vector
346 vector<string> theseNames;
347 m->splitAtComma(names, theseNames);
349 for (int i = 0; i < theseNames.size(); i++) { nameMap[theseNames[i]] = name; }
351 if (m->control_pressed) { inNames.close(); nameMap.clear(); return 0; }
357 catch(exception& e) {
358 m->errorOut(e, "ClassifyOtuCommand", "readNamesFile");
362 //**********************************************************************************************************************
363 int ClassifyOtuCommand::readTaxonomyFile() {
367 m->openInputFile(taxfile, in);
375 //are there confidence scores, if so remove them
376 if (tax.find_first_of('(') != -1) { m->removeConfidences(tax); }
380 if (m->control_pressed) { in.close(); taxMap.clear(); return 0; }
386 catch(exception& e) {
387 m->errorOut(e, "ClassifyOtuCommand", "readTaxonomyFile");
391 //**********************************************************************************************************************
392 vector<string> ClassifyOtuCommand::findConsensusTaxonomy(int bin, ListVector* thisList, int& size, string& conTax) {
395 vector<string> names;
396 vector<string> allNames;
397 map<string, string>::iterator it;
398 map<string, string>::iterator it2;
400 //parse names into vector
401 string binnames = thisList->get(bin);
402 m->splitAtComma(binnames, names);
404 //create a tree containing sequences from this bin
405 PhyloTree* phylo = new PhyloTree();
408 for (int i = 0; i < names.size(); i++) {
410 //if namesfile include the names
411 if (namefile != "") {
413 //is this sequence in the name file - namemap maps seqName -> repSeqName
414 it2 = nameMap.find(names[i]);
416 if (it2 == nameMap.end()) { //this name is not in name file, skip it
417 m->mothurOut(names[i] + " is not in your name file. I will not include it in the consensus."); m->mothurOutEndLine();
420 //is this sequence in the taxonomy file - look for repSeqName since we are assuming the taxonomy file is unique
421 it = taxMap.find(it2->second);
423 if (it == taxMap.end()) { //this name is not in taxonomy file, skip it
425 if (names[i] != it2->second) { m->mothurOut(names[i] + " is represented by " + it2->second + " and is not in your taxonomy file. I will not include it in the consensus."); m->mothurOutEndLine(); }
426 else { m->mothurOut(names[i] + " is not in your taxonomy file. I will not include it in the consensus."); m->mothurOutEndLine(); }
430 phylo->addSeqToTree(names[i], it->second);
432 allNames.push_back(names[i]);
437 //is this sequence in the taxonomy file - look for repSeqName since we are assuming the taxonomy file is unique
438 it = taxMap.find(names[i]);
440 if (it == taxMap.end()) { //this name is not in taxonomy file, skip it
441 m->mothurOut(names[i] + " is not in your taxonomy file. I will not include it in the consensus."); m->mothurOutEndLine();
444 phylo->addSeqToTree(names[i], it->second);
446 allNames.push_back(names[i]);
451 if (m->control_pressed) { delete phylo; return allNames; }
456 phylo->assignHeirarchyIDs(0);
458 TaxNode currentNode = phylo->get(0);
461 while (currentNode.children.size() != 0) { //you still have more to explore
464 int bestChildSize = 0;
466 //go through children
467 for (map<string, int>::iterator itChild = currentNode.children.begin(); itChild != currentNode.children.end(); itChild++) {
469 TaxNode temp = phylo->get(itChild->second);
471 //select child with largest accesions - most seqs assigned to it
472 if (temp.accessions.size() > bestChildSize) {
473 bestChild = phylo->get(itChild->second);
474 bestChildSize = temp.accessions.size();
479 //phylotree adds an extra unknown so we want to remove that
480 if (bestChild.name == "unknown") { bestChildSize--; }
482 //is this taxonomy above cutoff
483 int consensusConfidence = ceil((bestChildSize / (float) size) * 100);
485 if (consensusConfidence >= cutoff) { //if yes, add it
487 conTax += bestChild.name + "(" + toString(consensusConfidence) + ");";
489 conTax += bestChild.name + ";";
497 currentNode = bestChild;
500 if (myLevel != phylo->getMaxLevel()) {
501 while (myLevel != phylo->getMaxLevel()) {
502 conTax += "unclassified;";
506 if (conTax == "") { conTax = "no_consensus;"; }
513 catch(exception& e) {
514 m->errorOut(e, "ClassifyOtuCommand", "findConsensusTaxonomy");
519 //**********************************************************************************************************************
520 int ClassifyOtuCommand::process(ListVector* processList) {
526 if (outputDir == "") { outputDir += m->hasPath(listfile); }
529 string outputFile = outputDir + m->getRootName(m->getSimpleName(listfile)) + processList->getLabel() + ".cons.taxonomy";
530 m->openOutputFile(outputFile, out);
531 outputNames.push_back(outputFile); outputTypes["constaxonomy"].push_back(outputFile);
534 string outputSumFile = outputDir + m->getRootName(m->getSimpleName(listfile)) + processList->getLabel() + ".cons.tax.summary";
535 m->openOutputFile(outputSumFile, outSum);
536 outputNames.push_back(outputSumFile); outputTypes["taxsummary"].push_back(outputSumFile);
538 out << "OTU\tSize\tTaxonomy" << endl;
540 PhyloSummary* taxaSum;
541 if (refTaxonomy != "") {
542 taxaSum = new PhyloSummary(refTaxonomy, groupfile);
544 taxaSum = new PhyloSummary(groupfile);
547 //for each bin in the list vector
548 for (int i = 0; i < processList->getNumBins(); i++) {
550 if (m->control_pressed) { break; }
552 vector<string> names;
553 names = findConsensusTaxonomy(i, processList, size, conTax);
555 if (m->control_pressed) { out.close(); return 0; }
557 //output to new names file
558 out << (i+1) << '\t' << size << '\t' << conTax << endl;
560 string noConfidenceConTax = conTax;
561 m->removeConfidences(noConfidenceConTax);
563 //add this bins taxonomy to summary
564 if (basis == "sequence") {
565 for(int j = 0; j < names.size(); j++) { taxaSum->addSeqToTree(names[j], noConfidenceConTax); }
567 taxaSum->addSeqToTree(noConfidenceConTax, names);
574 taxaSum->print(outSum);
582 catch(exception& e) {
583 m->errorOut(e, "ClassifyOtuCommand", "process");
587 /**************************************************************************************************/
588 string ClassifyOtuCommand::addUnclassifieds(string tax, int maxlevel) {
590 string newTax, taxon;
593 //keep what you have counting the levels
594 while (tax.find_first_of(';') != -1) {
596 taxon = tax.substr(0,tax.find_first_of(';'))+';';
597 tax = tax.substr(tax.find_first_of(';')+1, tax.length());
602 //add "unclassified" until you reach maxLevel
603 while (level < maxlevel) {
604 newTax += "unclassified;";
610 catch(exception& e) {
611 m->errorOut(e, "ClassifyOtuCommand", "addUnclassifieds");
615 //**********************************************************************************************************************