2 * treegroupscommand.cpp
5 * Created by Sarah Westcott on 4/8/09.
6 * Copyright 2009 Schloss Lab UMASS Amherst. All rights reserved.
10 #include "treegroupscommand.h"
11 #include "sharedjabund.h"
12 #include "sharedsorabund.h"
13 #include "sharedjclass.h"
14 #include "sharedsorclass.h"
15 #include "sharedjest.h"
16 #include "sharedsorest.h"
17 #include "sharedthetayc.h"
18 #include "sharedthetan.h"
19 #include "sharedmorisitahorn.h"
20 #include "sharedbraycurtis.h"
23 //**********************************************************************************************************************
25 TreeGroupCommand::TreeGroupCommand(string option) {
27 globaldata = GlobalData::getInstance();
34 //allow user to run help
35 if(option == "help") { validCalculator = new ValidCalculators(); help(); abort = true; }
38 //valid paramters for this command
39 string Array[] = {"label","calc","groups", "phylip", "column", "name", "precision","cutoff","outputdir","inputdir"};
40 vector<string> myArray (Array, Array+(sizeof(Array)/sizeof(string)));
42 OptionParser parser(option);
43 map<string, string> parameters = parser. getParameters();
45 ValidParameters validParameter;
46 map<string, string>::iterator it;
48 //check to make sure all parameters are valid for command
49 for (it = parameters.begin(); it != parameters.end(); it++) {
50 if (validParameter.isValidParameter(it->first, myArray, it->second) != true) { abort = true; }
53 //if the user changes the input directory command factory will send this info to us in the output parameter
54 string inputDir = validParameter.validFile(parameters, "inputdir", false);
55 if (inputDir == "not found"){ inputDir = ""; }
58 it = parameters.find("phylip");
59 //user has given a template file
60 if(it != parameters.end()){
61 path = hasPath(it->second);
62 //if the user has not given a path then, add inputdir. else leave path alone.
63 if (path == "") { parameters["phylip"] = inputDir + it->second; }
66 it = parameters.find("column");
67 //user has given a template file
68 if(it != parameters.end()){
69 path = hasPath(it->second);
70 //if the user has not given a path then, add inputdir. else leave path alone.
71 if (path == "") { parameters["column"] = inputDir + it->second; }
74 it = parameters.find("name");
75 //user has given a template file
76 if(it != parameters.end()){
77 path = hasPath(it->second);
78 //if the user has not given a path then, add inputdir. else leave path alone.
79 if (path == "") { parameters["name"] = inputDir + it->second; }
83 format = globaldata->getFormat();
86 phylipfile = validParameter.validFile(parameters, "phylip", true);
87 if (phylipfile == "not open") { abort = true; }
88 else if (phylipfile == "not found") { phylipfile = ""; }
89 else { format = "phylip"; globaldata->setPhylipFile(phylipfile); }
91 columnfile = validParameter.validFile(parameters, "column", true);
92 if (columnfile == "not open") { abort = true; }
93 else if (columnfile == "not found") { columnfile = ""; }
94 else { format = "column"; globaldata->setColumnFile(columnfile); }
96 namefile = validParameter.validFile(parameters, "name", true);
97 if (namefile == "not open") { abort = true; }
98 else if (namefile == "not found") { namefile = ""; }
99 else { globaldata->setNameFile(namefile); }
101 //error checking on files
102 if ((globaldata->getSharedFile() == "") && ((phylipfile == "") && (columnfile == ""))) { m->mothurOut("You must run the read.otu command or provide a distance file before running the tree.shared command."); m->mothurOutEndLine(); abort = true; }
103 else if ((phylipfile != "") && (columnfile != "")) { m->mothurOut("When running the tree.shared command with a distance file you may not use both the column and the phylip parameters."); m->mothurOutEndLine(); abort = true; }
105 if (columnfile != "") {
106 if (namefile == "") { m->mothurOut("You need to provide a namefile if you are going to use the column format."); m->mothurOutEndLine(); abort = true; }
109 //check for optional parameter and set defaults
110 // ...at some point should added some additional type checking...
111 label = validParameter.validFile(parameters, "label", false);
112 if (label == "not found") { label = ""; }
114 if(label != "all") { splitAtDash(label, labels); allLines = 0; }
115 else { allLines = 1; }
118 //if the user has not specified any labels use the ones from read.otu
120 allLines = globaldata->allLines;
121 labels = globaldata->labels;
124 groups = validParameter.validFile(parameters, "groups", false);
125 if (groups == "not found") { groups = ""; }
127 splitAtDash(groups, Groups);
128 globaldata->Groups = Groups;
131 calc = validParameter.validFile(parameters, "calc", false);
132 if (calc == "not found") { calc = "jclass-thetayc"; }
134 if (calc == "default") { calc = "jclass-thetayc"; }
136 splitAtDash(calc, Estimators);
139 temp = validParameter.validFile(parameters, "precision", false); if (temp == "not found") { temp = "100"; }
140 convert(temp, precision);
142 temp = validParameter.validFile(parameters, "cutoff", false); if (temp == "not found") { temp = "10"; }
143 convert(temp, cutoff);
144 cutoff += (5 / (precision * 10.0));
146 //if the user changes the output directory command factory will send this info to us in the output parameter
147 outputDir = validParameter.validFile(parameters, "outputdir", false); if (outputDir == "not found"){
149 outputDir += hasPath(globaldata->inputFileName); //if user entered a file with a path then preserve it
153 if (abort == false) {
155 validCalculator = new ValidCalculators();
157 if (format == "sharedfile") {
159 for (i=0; i<Estimators.size(); i++) {
160 if (validCalculator->isValidCalculator("treegroup", Estimators[i]) == true) {
161 if (Estimators[i] == "jabund") {
162 treeCalculators.push_back(new JAbund());
163 }else if (Estimators[i] == "sorabund") {
164 treeCalculators.push_back(new SorAbund());
165 }else if (Estimators[i] == "jclass") {
166 treeCalculators.push_back(new Jclass());
167 }else if (Estimators[i] == "sorclass") {
168 treeCalculators.push_back(new SorClass());
169 }else if (Estimators[i] == "jest") {
170 treeCalculators.push_back(new Jest());
171 }else if (Estimators[i] == "sorest") {
172 treeCalculators.push_back(new SorEst());
173 }else if (Estimators[i] == "thetayc") {
174 treeCalculators.push_back(new ThetaYC());
175 }else if (Estimators[i] == "thetan") {
176 treeCalculators.push_back(new ThetaN());
177 }else if (Estimators[i] == "morisitahorn") {
178 treeCalculators.push_back(new MorHorn());
179 }else if (Estimators[i] == "braycurtis") {
180 treeCalculators.push_back(new BrayCurtis());
189 catch(exception& e) {
190 m->errorOut(e, "TreeGroupCommand", "TreeGroupCommand");
195 //**********************************************************************************************************************
197 void TreeGroupCommand::help(){
199 m->mothurOut("The tree.shared command creates a .tre to represent the similiarity between groups or sequences.\n");
200 m->mothurOut("The tree.shared command can only be executed after a successful read.otu command or by providing a distance file.\n");
201 m->mothurOut("The tree.shared command parameters are groups, calc, phylip, column, name, cutoff, precision and label.\n");
202 m->mothurOut("The groups parameter allows you to specify which of the groups in your groupfile you would like included used.\n");
203 m->mothurOut("The group names are separated by dashes. The label allow you to select what distance levels you would like trees created for, and are also separated by dashes.\n");
204 m->mothurOut("The phylip or column parameter are required if you do not run the read.otu command first, and only one may be used. If you use a column file the name filename is required. \n");
205 m->mothurOut("If you do not provide a cutoff value 10.00 is assumed. If you do not provide a precision value then 100 is assumed.\n");
206 m->mothurOut("The tree.shared command should be in the following format: tree.shared(groups=yourGroups, calc=yourCalcs, label=yourLabels).\n");
207 m->mothurOut("Example tree.shared(groups=A-B-C, calc=jabund-sorabund).\n");
208 m->mothurOut("The default value for groups is all the groups in your groupfile.\n");
209 m->mothurOut("The default value for calc is jclass-thetayc.\n");
210 m->mothurOut("The tree.shared command outputs a .tre file for each calculator you specify at each distance you choose.\n");
211 validCalculator->printCalc("treegroup", cout);
212 m->mothurOut("Or the tree.shared command can be in the following format: tree.shared(phylip=yourPhylipFile).\n");
213 m->mothurOut("Example tree.shared(phylip=abrecovery.dist).\n");
214 m->mothurOut("Note: No spaces between parameter labels (i.e. groups), '=' and parameters (i.e.yourGroups).\n\n");
216 catch(exception& e) {
217 m->errorOut(e, "TreeGroupCommand", "help");
223 //**********************************************************************************************************************
225 TreeGroupCommand::~TreeGroupCommand(){
226 if (abort == false) {
228 if (format == "sharedfile") { delete read; delete input; globaldata->ginput = NULL; }
229 else { delete readMatrix; delete matrix; delete list; }
230 delete tmap; globaldata->gTreemap = NULL;
231 delete validCalculator;
236 //**********************************************************************************************************************
238 int TreeGroupCommand::execute(){
241 if (abort == true) { return 0; }
243 if (format == "sharedfile") {
244 //if the users entered no valid calculators don't execute command
245 if (treeCalculators.size() == 0) { m->mothurOut("You have given no valid calculators."); m->mothurOutEndLine(); return 0; }
248 read = new ReadOTUFile(globaldata->inputFileName);
249 read->read(&*globaldata);
251 input = globaldata->ginput;
252 lookup = input->getSharedRAbundVectors();
253 lastLabel = lookup[0]->getLabel();
255 if (lookup.size() < 2) { m->mothurOut("You have not provided enough valid groups. I cannot run the command."); m->mothurOutEndLine(); return 0; }
257 //used in tree constructor
258 globaldata->runParse = false;
264 filename = globaldata->inputFileName;
266 if (format == "column") { readMatrix = new ReadColumnMatrix(filename); }
267 else if (format == "phylip") { readMatrix = new ReadPhylipMatrix(filename); }
269 readMatrix->setCutoff(cutoff);
272 nameMap = new NameAssignment(namefile);
279 readMatrix->read(nameMap);
280 list = readMatrix->getListVector();
281 matrix = readMatrix->getMatrix();
284 tmap = new TreeMap();
286 globaldata->gTreemap = tmap;
288 globaldata->Groups = tmap->namesOfGroups;
290 //clear globaldatas old tree names if any
291 globaldata->Treenames.clear();
293 //fills globaldatas tree names
294 globaldata->Treenames = globaldata->Groups;
296 //used in tree constructor
297 globaldata->runParse = false;
301 //create a new filename
302 outputFile = outputDir + getRootName(getSimpleName(globaldata->inputFileName)) + "tre";
303 outputNames.push_back(outputFile);
306 m->mothurOut("Tree complete. "); m->mothurOutEndLine();
309 //reset groups parameter
310 globaldata->Groups.clear();
312 m->mothurOutEndLine();
313 m->mothurOut("Output File Names: "); m->mothurOutEndLine();
314 for (int i = 0; i < outputNames.size(); i++) { m->mothurOut(outputNames[i]); m->mothurOutEndLine(); }
315 m->mothurOutEndLine();
319 catch(exception& e) {
320 m->errorOut(e, "TreeGroupCommand", "execute");
324 //**********************************************************************************************************************
326 void TreeGroupCommand::createTree(){
331 //do merges and create tree structure by setting parents and children
332 //there are numGroups - 1 merges to do
333 for (int i = 0; i < (numGroups - 1); i++) {
334 float largest = -1000.0;
337 //find largest value in sims matrix by searching lower triangle
338 for (int j = 1; j < simMatrix.size(); j++) {
339 for (int k = 0; k < j; k++) {
340 if (simMatrix[j][k] > largest) { largest = simMatrix[j][k]; row = j; column = k; }
344 //set non-leaf node info and update leaves to know their parents
346 t->tree[numGroups + i].setChildren(index[row], index[column]);
349 t->tree[index[row]].setParent(numGroups + i);
350 t->tree[index[column]].setParent(numGroups + i);
352 //blength = distance / 2;
353 float blength = ((1.0 - largest) / 2);
356 t->tree[index[row]].setBranchLength(blength - t->tree[index[row]].getLengthToLeaves());
357 t->tree[index[column]].setBranchLength(blength - t->tree[index[column]].getLengthToLeaves());
359 //set your length to leaves to your childs length plus branchlength
360 t->tree[numGroups + i].setLengthToLeaves(t->tree[index[row]].getLengthToLeaves() + t->tree[index[row]].getBranchLength());
364 index[row] = numGroups+i;
365 index[column] = numGroups+i;
367 //remove highest value that caused the merge.
368 simMatrix[row][column] = -1000.0;
369 simMatrix[column][row] = -1000.0;
371 //merge values in simsMatrix
372 for (int n = 0; n < simMatrix.size(); n++) {
373 //row becomes merge of 2 groups
374 simMatrix[row][n] = (simMatrix[row][n] + simMatrix[column][n]) / 2;
375 simMatrix[n][row] = simMatrix[row][n];
377 simMatrix[column][n] = -1000.0;
378 simMatrix[n][column] = -1000.0;
382 //adjust tree to make sure root to tip length is .5
383 int root = t->findRoot();
384 t->tree[root].setBranchLength((0.5 - t->tree[root].getLengthToLeaves()));
390 t->createNewickFile(outputFile);
396 catch(exception& e) {
397 m->errorOut(e, "TreeGroupCommand", "createTree");
401 /***********************************************************/
402 void TreeGroupCommand::printSims(ostream& out) {
405 //output column headers
407 //for (int i = 0; i < lookup.size(); i++) { out << lookup[i]->getGroup() << '\t'; }
411 for (int m = 0; m < simMatrix.size(); m++) {
412 //out << lookup[m]->getGroup() << '\t';
413 for (int n = 0; n < simMatrix.size(); n++) {
414 out << simMatrix[m][n] << '\t';
420 catch(exception& e) {
421 m->errorOut(e, "TreeGroupCommand", "printSims");
425 /***********************************************************/
426 void TreeGroupCommand::makeSimsDist() {
428 numGroups = list->size();
432 for (int g = 0; g < numGroups; g++) { index[g] = g; }
434 //initialize simMatrix
436 simMatrix.resize(numGroups);
437 for (int m = 0; m < simMatrix.size(); m++) {
438 for (int j = 0; j < simMatrix.size(); j++) {
439 simMatrix[m].push_back(0.0);
443 //go through sparse matrix and fill sims
444 //go through each cell in the sparsematrix
445 for(MatData currentCell = matrix->begin(); currentCell != matrix->end(); currentCell++){
446 //similairity = -(distance-1)
447 simMatrix[currentCell->row][currentCell->column] = -(currentCell->dist -1.0);
448 simMatrix[currentCell->column][currentCell->row] = -(currentCell->dist -1.0);
453 catch(exception& e) {
454 m->errorOut(e, "TreeGroupCommand", "makeSimsDist");
459 /***********************************************************/
460 void TreeGroupCommand::makeSimsShared() {
463 //clear globaldatas old tree names if any
464 globaldata->Treenames.clear();
466 //fills globaldatas tree names
467 globaldata->Treenames = globaldata->Groups;
469 //create treemap class from groupmap for tree class to use
470 tmap = new TreeMap();
471 tmap->makeSim(globaldata->gGroupmap);
472 globaldata->gTreemap = tmap;
474 set<string> processedLabels;
475 set<string> userLabels = labels;
477 //as long as you are not at the end of the file or done wih the lines you want
478 while((lookup[0] != NULL) && ((allLines == 1) || (userLabels.size() != 0))) {
480 if(allLines == 1 || labels.count(lookup[0]->getLabel()) == 1){
481 m->mothurOut(lookup[0]->getLabel()); m->mothurOutEndLine();
484 processedLabels.insert(lookup[0]->getLabel());
485 userLabels.erase(lookup[0]->getLabel());
488 if ((anyLabelsToProcess(lookup[0]->getLabel(), userLabels, "") == true) && (processedLabels.count(lastLabel) != 1)) {
489 string saveLabel = lookup[0]->getLabel();
491 for (int i = 0; i < lookup.size(); i++) { delete lookup[i]; }
492 lookup = input->getSharedRAbundVectors(lastLabel);
494 m->mothurOut(lookup[0]->getLabel()); m->mothurOutEndLine();
497 processedLabels.insert(lookup[0]->getLabel());
498 userLabels.erase(lookup[0]->getLabel());
500 //restore real lastlabel to save below
501 lookup[0]->setLabel(saveLabel);
504 lastLabel = lookup[0]->getLabel();
506 //get next line to process
507 for (int i = 0; i < lookup.size(); i++) { delete lookup[i]; }
508 lookup = input->getSharedRAbundVectors();
511 //output error messages about any remaining user labels
512 set<string>::iterator it;
513 bool needToRun = false;
514 for (it = userLabels.begin(); it != userLabels.end(); it++) {
515 m->mothurOut("Your file does not include the label " + *it);
516 if (processedLabels.count(lastLabel) != 1) {
517 m->mothurOut(". I will use " + lastLabel + "."); m->mothurOutEndLine();
520 m->mothurOut(". Please refer to " + lastLabel + "."); m->mothurOutEndLine();
524 //run last label if you need to
525 if (needToRun == true) {
526 for (int i = 0; i < lookup.size(); i++) { if (lookup[i] != NULL) { delete lookup[i]; } }
527 lookup = input->getSharedRAbundVectors(lastLabel);
529 m->mothurOut(lookup[0]->getLabel()); m->mothurOutEndLine();
531 for (int i = 0; i < lookup.size(); i++) { delete lookup[i]; }
534 for(int i = 0 ; i < treeCalculators.size(); i++) { delete treeCalculators[i]; }
536 catch(exception& e) {
537 m->errorOut(e, "TreeGroupCommand", "makeSimsShared");
542 /***********************************************************/
543 void TreeGroupCommand::process(vector<SharedRAbundVector*> thisLookup) {
546 vector<SharedRAbundVector*> subset;
547 numGroups = thisLookup.size();
549 //for each calculator
550 for(int i = 0 ; i < treeCalculators.size(); i++) {
551 //initialize simMatrix
553 simMatrix.resize(numGroups);
554 for (int m = 0; m < simMatrix.size(); m++) {
555 for (int j = 0; j < simMatrix.size(); j++) {
556 simMatrix[m].push_back(0.0);
562 for (int g = 0; g < numGroups; g++) { index[g] = g; }
564 //create a new filename
565 outputFile = outputDir + getRootName(getSimpleName(globaldata->inputFileName)) + treeCalculators[i]->getName() + "." + thisLookup[0]->getLabel() + ".tre";
566 outputNames.push_back(outputFile);
568 for (int k = 0; k < thisLookup.size(); k++) {
569 for (int l = k; l < thisLookup.size(); l++) {
570 if (k != l) { //we dont need to similiarity of a groups to itself
571 //get estimated similarity between 2 groups
573 subset.clear(); //clear out old pair of sharedrabunds
574 //add new pair of sharedrabunds
575 subset.push_back(thisLookup[k]); subset.push_back(thisLookup[l]);
577 data = treeCalculators[i]->getValues(subset); //saves the calculator outputs
578 //save values in similarity matrix
579 simMatrix[k][l] = data[0];
580 simMatrix[l][k] = data[0];
585 //creates tree from similarity matrix and write out file
590 catch(exception& e) {
591 m->errorOut(e, "TreeGroupCommand", "process");
595 /***********************************************************/