2 * preclustercommand.cpp
5 * Created by westcott on 12/21/09.
6 * Copyright 2009 Schloss Lab. All rights reserved.
10 #include "preclustercommand.h"
11 #include "deconvolutecommand.h"
13 //**********************************************************************************************************************
14 vector<string> PreClusterCommand::setParameters(){
16 CommandParameter pfasta("fasta", "InputTypes", "", "", "none", "none", "none","fasta-name",false,true,true); parameters.push_back(pfasta);
17 CommandParameter pname("name", "InputTypes", "", "", "NameCount", "none", "none","name",false,false,true); parameters.push_back(pname);
18 CommandParameter pcount("count", "InputTypes", "", "", "NameCount-CountGroup", "none", "none","count",false,false,true); parameters.push_back(pcount);
19 CommandParameter pgroup("group", "InputTypes", "", "", "CountGroup", "none", "none","",false,false,true); parameters.push_back(pgroup);
20 CommandParameter pdiffs("diffs", "Number", "", "1", "", "", "","",false,false,true); parameters.push_back(pdiffs);
21 CommandParameter pprocessors("processors", "Number", "", "1", "", "", "","",false,false,true); parameters.push_back(pprocessors);
22 CommandParameter ptopdown("topdown", "Boolean", "", "T", "", "", "","",false,false); parameters.push_back(ptopdown);
24 CommandParameter pinputdir("inputdir", "String", "", "", "", "", "","",false,false); parameters.push_back(pinputdir);
25 CommandParameter poutputdir("outputdir", "String", "", "", "", "", "","",false,false); parameters.push_back(poutputdir);
27 vector<string> myArray;
28 for (int i = 0; i < parameters.size(); i++) { myArray.push_back(parameters[i].name); }
32 m->errorOut(e, "PreClusterCommand", "setParameters");
36 //**********************************************************************************************************************
37 string PreClusterCommand::getHelpString(){
39 string helpString = "";
40 helpString += "The pre.cluster command groups sequences that are within a given number of base mismatches.\n";
41 helpString += "The pre.cluster command outputs a new fasta and name file.\n";
42 helpString += "The pre.cluster command parameters are fasta, name, group, count, topdown, processors and diffs. The fasta parameter is required. \n";
43 helpString += "The name parameter allows you to give a list of seqs that are identical. This file is 2 columns, first column is name or representative sequence, second column is a list of its identical sequences separated by commas.\n";
44 helpString += "The group parameter allows you to provide a group file so you can cluster by group. \n";
45 helpString += "The count parameter allows you to provide a count file so you can cluster by group. \n";
46 helpString += "The diffs parameter allows you to specify maximum number of mismatched bases allowed between sequences in a grouping. The default is 1.\n";
47 helpString += "The topdown parameter allows you to specify whether to cluster from largest abundance to smallest or smallest to largest. Default=T, meaning largest to smallest.\n";
48 helpString += "The pre.cluster command should be in the following format: \n";
49 helpString += "pre.cluster(fasta=yourFastaFile, names=yourNamesFile, diffs=yourMaxDiffs) \n";
50 helpString += "Example pre.cluster(fasta=amazon.fasta, diffs=2).\n";
51 helpString += "Note: No spaces between parameter labels (i.e. fasta), '=' and parameters (i.e.yourFasta).\n";
55 m->errorOut(e, "PreClusterCommand", "getHelpString");
59 //**********************************************************************************************************************
60 string PreClusterCommand::getOutputPattern(string type) {
64 if (type == "fasta") { pattern = "[filename],precluster,[extension]"; }
65 else if (type == "name") { pattern = "[filename],precluster.names"; }
66 else if (type == "count") { pattern = "[filename],precluster.count_table"; }
67 else if (type == "map") { pattern = "[filename],precluster.map"; }
68 else { m->mothurOut("[ERROR]: No definition for type " + type + " output pattern.\n"); m->control_pressed = true; }
73 m->errorOut(e, "PreClusterCommand", "getOutputPattern");
77 //**********************************************************************************************************************
78 PreClusterCommand::PreClusterCommand(){
80 abort = true; calledHelp = true;
82 vector<string> tempOutNames;
83 outputTypes["fasta"] = tempOutNames;
84 outputTypes["name"] = tempOutNames;
85 outputTypes["count"] = tempOutNames;
86 outputTypes["map"] = tempOutNames;
89 m->errorOut(e, "PreClusterCommand", "PreClusterCommand");
93 //**********************************************************************************************************************
95 PreClusterCommand::PreClusterCommand(string option) {
97 abort = false; calledHelp = false;
99 //allow user to run help
100 if(option == "help") { help(); abort = true; calledHelp = true; }
101 else if(option == "citation") { citation(); abort = true; calledHelp = true;}
104 vector<string> myArray = setParameters();
106 OptionParser parser(option);
107 map<string, string> parameters = parser.getParameters();
109 ValidParameters validParameter;
110 map<string, string>::iterator it;
112 //check to make sure all parameters are valid for command
113 for (map<string, string>::iterator it2 = parameters.begin(); it2 != parameters.end(); it2++) {
114 if (validParameter.isValidParameter(it2->first, myArray, it2->second) != true) { abort = true; }
117 //initialize outputTypes
118 vector<string> tempOutNames;
119 outputTypes["fasta"] = tempOutNames;
120 outputTypes["name"] = tempOutNames;
121 outputTypes["map"] = tempOutNames;
122 outputTypes["count"] = tempOutNames;
124 //if the user changes the input directory command factory will send this info to us in the output parameter
125 string inputDir = validParameter.validFile(parameters, "inputdir", false);
126 if (inputDir == "not found"){ inputDir = ""; }
129 it = parameters.find("fasta");
130 //user has given a template file
131 if(it != parameters.end()){
132 path = m->hasPath(it->second);
133 //if the user has not given a path then, add inputdir. else leave path alone.
134 if (path == "") { parameters["fasta"] = inputDir + it->second; }
137 it = parameters.find("name");
138 //user has given a template file
139 if(it != parameters.end()){
140 path = m->hasPath(it->second);
141 //if the user has not given a path then, add inputdir. else leave path alone.
142 if (path == "") { parameters["name"] = inputDir + it->second; }
145 it = parameters.find("group");
146 //user has given a template file
147 if(it != parameters.end()){
148 path = m->hasPath(it->second);
149 //if the user has not given a path then, add inputdir. else leave path alone.
150 if (path == "") { parameters["group"] = inputDir + it->second; }
153 it = parameters.find("count");
154 //user has given a template file
155 if(it != parameters.end()){
156 path = m->hasPath(it->second);
157 //if the user has not given a path then, add inputdir. else leave path alone.
158 if (path == "") { parameters["count"] = inputDir + it->second; }
162 //check for required parameters
163 fastafile = validParameter.validFile(parameters, "fasta", true);
164 if (fastafile == "not found") {
165 fastafile = m->getFastaFile();
166 if (fastafile != "") { m->mothurOut("Using " + fastafile + " as input file for the fasta parameter."); m->mothurOutEndLine(); }
167 else { m->mothurOut("You have no current fastafile and the fasta parameter is required."); m->mothurOutEndLine(); abort = true; }
169 else if (fastafile == "not open") { abort = true; }
170 else { m->setFastaFile(fastafile); }
172 //if the user changes the output directory command factory will send this info to us in the output parameter
173 outputDir = validParameter.validFile(parameters, "outputdir", false); if (outputDir == "not found"){
175 outputDir += m->hasPath(fastafile); //if user entered a file with a path then preserve it
178 //check for optional parameter and set defaults
179 // ...at some point should added some additional type checking...
180 namefile = validParameter.validFile(parameters, "name", true);
181 if (namefile == "not found") { namefile = ""; }
182 else if (namefile == "not open") { namefile = ""; abort = true; }
183 else { m->setNameFile(namefile); }
185 groupfile = validParameter.validFile(parameters, "group", true);
186 if (groupfile == "not found") { groupfile = ""; bygroup = false; }
187 else if (groupfile == "not open") { abort = true; groupfile = ""; }
188 else { m->setGroupFile(groupfile); bygroup = true; }
190 countfile = validParameter.validFile(parameters, "count", true);
191 if (countfile == "not found") { countfile = ""; }
192 else if (countfile == "not open") { abort = true; countfile = ""; }
194 m->setCountTableFile(countfile);
195 ct.readTable(countfile, true, false);
196 if (ct.hasGroupInfo()) { bygroup = true; }
197 else { bygroup = false; }
200 if ((namefile != "") && (countfile != "")) {
201 m->mothurOut("[ERROR]: you may only use one of the following: name or count."); m->mothurOutEndLine(); abort = true;
204 if ((groupfile != "") && (countfile != "")) {
205 m->mothurOut("[ERROR]: you may only use one of the following: group or count."); m->mothurOutEndLine(); abort=true;
209 string temp = validParameter.validFile(parameters, "diffs", false); if(temp == "not found"){ temp = "1"; }
210 m->mothurConvert(temp, diffs);
212 temp = validParameter.validFile(parameters, "processors", false); if (temp == "not found"){ temp = m->getProcessors(); }
213 m->setProcessors(temp);
214 m->mothurConvert(temp, processors);
216 temp = validParameter.validFile(parameters, "topdown", false); if(temp == "not found"){ temp = "T"; }
217 topdown = m->isTrue(temp);
219 if (countfile == "") {
220 if (namefile == "") {
221 vector<string> files; files.push_back(fastafile);
222 parser.getNameFile(files);
228 catch(exception& e) {
229 m->errorOut(e, "PreClusterCommand", "PreClusterCommand");
233 //**********************************************************************************************************************
235 int PreClusterCommand::execute(){
238 if (abort == true) { if (calledHelp) { return 0; } return 2; }
240 int start = time(NULL);
242 string fileroot = outputDir + m->getRootName(m->getSimpleName(fastafile));
243 map<string, string> variables;
244 variables["[filename]"] = fileroot;
245 string newNamesFile = getOutputFileName("name",variables);
246 string newCountFile = getOutputFileName("count",variables);
247 string newMapFile = getOutputFileName("map",variables); //add group name if by group
248 variables["[extension]"] = m->getExtension(fastafile);
249 string newFastaFile = getOutputFileName("fasta", variables);
250 outputNames.push_back(newFastaFile); outputTypes["fasta"].push_back(newFastaFile);
251 if (countfile == "") { outputNames.push_back(newNamesFile); outputTypes["name"].push_back(newNamesFile); }
252 else { outputNames.push_back(newCountFile); outputTypes["count"].push_back(newCountFile); }
255 //clear out old files
256 ofstream outFasta; m->openOutputFile(newFastaFile, outFasta); outFasta.close();
257 ofstream outNames; m->openOutputFile(newNamesFile, outNames); outNames.close();
258 newMapFile = fileroot + "precluster.";
260 //parse fasta and name file by group
261 vector<string> groups;
262 if (countfile != "") {
263 cparser = new SequenceCountParser(countfile, fastafile);
264 groups = cparser->getNamesOfGroups();
266 if (namefile != "") { parser = new SequenceParser(groupfile, fastafile, namefile); }
267 else { parser = new SequenceParser(groupfile, fastafile); }
268 groups = parser->getNamesOfGroups();
271 if(processors == 1) { driverGroups(newFastaFile, newNamesFile, newMapFile, 0, groups.size(), groups); }
272 else { createProcessesGroups(newFastaFile, newNamesFile, newMapFile, groups); }
274 if (countfile != "") {
275 mergeGroupCounts(newCountFile, newNamesFile, newFastaFile);
279 //run unique.seqs for deconvolute results
280 string inputString = "fasta=" + newFastaFile;
281 if (namefile != "") { inputString += ", name=" + newNamesFile; }
282 m->mothurOutEndLine();
283 m->mothurOut("/******************************************/"); m->mothurOutEndLine();
284 m->mothurOut("Running command: unique.seqs(" + inputString + ")"); m->mothurOutEndLine();
285 m->mothurCalling = true;
287 Command* uniqueCommand = new DeconvoluteCommand(inputString);
288 uniqueCommand->execute();
290 map<string, vector<string> > filenames = uniqueCommand->getOutputFiles();
292 delete uniqueCommand;
293 m->mothurCalling = false;
294 m->mothurOut("/******************************************/"); m->mothurOutEndLine();
296 m->renameFile(filenames["fasta"][0], newFastaFile);
297 m->renameFile(filenames["name"][0], newNamesFile);
299 if (m->control_pressed) { for (int i = 0; i < outputNames.size(); i++) { m->mothurRemove(outputNames[i]); } return 0; }
300 m->mothurOut("It took " + toString(time(NULL) - start) + " secs to run pre.cluster."); m->mothurOutEndLine();
303 if (processors != 1) { m->mothurOut("When using running without group information mothur can only use 1 processor, continuing."); m->mothurOutEndLine(); processors = 1; }
304 if (namefile != "") { readNameFile(); }
306 //reads fasta file and return number of seqs
307 int numSeqs = readFASTA(); //fills alignSeqs and makes all seqs active
309 if (m->control_pressed) { for (int i = 0; i < outputNames.size(); i++) { m->mothurRemove(outputNames[i]); } return 0; }
311 if (numSeqs == 0) { m->mothurOut("Error reading fasta file...please correct."); m->mothurOutEndLine(); return 0; }
312 if (diffs > length) { m->mothurOut("Error: diffs is greater than your sequence length."); m->mothurOutEndLine(); return 0; }
314 int count = process(newMapFile);
315 outputNames.push_back(newMapFile); outputTypes["map"].push_back(newMapFile);
317 if (m->control_pressed) { for (int i = 0; i < outputNames.size(); i++) { m->mothurRemove(outputNames[i]); } return 0; }
319 m->mothurOut("Total number of sequences before precluster was " + toString(alignSeqs.size()) + "."); m->mothurOutEndLine();
320 m->mothurOut("pre.cluster removed " + toString(count) + " sequences."); m->mothurOutEndLine(); m->mothurOutEndLine();
321 if (countfile != "") { newNamesFile = newCountFile; }
322 printData(newFastaFile, newNamesFile, "");
324 m->mothurOut("It took " + toString(time(NULL) - start) + " secs to cluster " + toString(numSeqs) + " sequences."); m->mothurOutEndLine();
327 if (m->control_pressed) { for (int i = 0; i < outputNames.size(); i++) { m->mothurRemove(outputNames[i]); } return 0; }
329 m->mothurOutEndLine();
330 m->mothurOut("Output File Names: "); m->mothurOutEndLine();
331 for (int i = 0; i < outputNames.size(); i++) { m->mothurOut(outputNames[i]); m->mothurOutEndLine(); }
332 m->mothurOutEndLine();
334 //set fasta file as new current fastafile
336 itTypes = outputTypes.find("fasta");
337 if (itTypes != outputTypes.end()) {
338 if ((itTypes->second).size() != 0) { current = (itTypes->second)[0]; m->setFastaFile(current); }
341 itTypes = outputTypes.find("name");
342 if (itTypes != outputTypes.end()) {
343 if ((itTypes->second).size() != 0) { current = (itTypes->second)[0]; m->setNameFile(current); }
346 itTypes = outputTypes.find("count");
347 if (itTypes != outputTypes.end()) {
348 if ((itTypes->second).size() != 0) { current = (itTypes->second)[0]; m->setCountTableFile(current); }
354 catch(exception& e) {
355 m->errorOut(e, "PreClusterCommand", "execute");
359 /**************************************************************************************************/
360 int PreClusterCommand::createProcessesGroups(string newFName, string newNName, string newMFile, vector<string> groups) {
363 vector<int> processIDS;
369 if (groups.size() < processors) { processors = groups.size(); }
371 //divide the groups between the processors
372 vector<linePair> lines;
373 int remainingPairs = groups.size();
375 for (int remainingProcessors = processors; remainingProcessors > 0; remainingProcessors--) {
376 int numPairs = remainingPairs; //case for last processor
377 if (remainingProcessors != 1) { numPairs = ceil(remainingPairs / remainingProcessors); }
378 lines.push_back(linePair(startIndex, (startIndex+numPairs))); //startIndex, endIndex
379 startIndex = startIndex + numPairs;
380 remainingPairs = remainingPairs - numPairs;
383 #if defined (__APPLE__) || (__MACH__) || (linux) || (__linux) || (__linux__) || (__unix__) || (__unix)
385 //loop through and create all the processes you want
386 while (process != processors) {
390 processIDS.push_back(pid); //create map from line number to pid so you can append files in correct order later
394 num = driverGroups(newFName + m->mothurGetpid(process) + ".temp", newNName + m->mothurGetpid(process) + ".temp", newMFile, lines[process].start, lines[process].end, groups);
396 string tempFile = m->mothurGetpid(process) + ".outputNames.temp";
398 m->openOutputFile(tempFile, outTemp);
400 outTemp << outputNames.size();
401 for (int i = 0; i < outputNames.size(); i++) { outTemp << outputNames[i] << endl; }
406 m->mothurOut("[ERROR]: unable to spawn the number of processes you requested, reducing number to " + toString(process) + "\n"); processors = process;
407 for (int i = 0; i < processIDS.size(); i++) { kill (processIDS[i], SIGINT); }
416 processIDS.resize(0);
419 int remainingPairs = groups.size();
421 for (int remainingProcessors = processors; remainingProcessors > 0; remainingProcessors--) {
422 int numPairs = remainingPairs; //case for last processor
423 if (remainingProcessors != 1) { numPairs = ceil(remainingPairs / remainingProcessors); }
424 lines.push_back(linePair(startIndex, (startIndex+numPairs))); //startIndex, endIndex
425 startIndex = startIndex + numPairs;
426 remainingPairs = remainingPairs - numPairs;
429 while (process != processors) {
433 processIDS.push_back(pid); //create map from line number to pid so you can append files in correct order later
437 num = driverGroups(newFName + m->mothurGetpid(process) + ".temp", newNName + m->mothurGetpid(process) + ".temp", newMFile, lines[process].start, lines[process].end, groups);
439 string tempFile = m->mothurGetpid(process) + ".outputNames.temp";
441 m->openOutputFile(tempFile, outTemp);
443 outTemp << outputNames.size();
444 for (int i = 0; i < outputNames.size(); i++) { outTemp << outputNames[i] << endl; }
449 m->mothurOut("[ERROR]: unable to spawn the necessary processes."); m->mothurOutEndLine();
450 for (int i = 0; i < processIDS.size(); i++) { kill (processIDS[i], SIGINT); }
458 num = driverGroups(newFName, newNName, newMFile, lines[0].start, lines[0].end, groups);
460 //force parent to wait until all the processes are done
461 for (int i=0;i<processIDS.size();i++) {
462 int temp = processIDS[i];
466 for (int i = 0; i < processIDS.size(); i++) {
467 string tempFile = toString(processIDS[i]) + ".outputNames.temp";
469 m->openInputFile(tempFile, intemp);
473 for (int k = 0; k < num; k++) {
475 intemp >> name; m->gobble(intemp);
477 outputNames.push_back(name); outputTypes["map"].push_back(name);
480 m->mothurRemove(tempFile);
484 //////////////////////////////////////////////////////////////////////////////////////////////////////
485 //Windows version shared memory, so be careful when passing variables through the preClusterData struct.
486 //Above fork() will clone, so memory is separate, but that's not the case with windows,
487 //////////////////////////////////////////////////////////////////////////////////////////////////////
489 vector<preClusterData*> pDataArray;
490 DWORD dwThreadIdArray[processors-1];
491 HANDLE hThreadArray[processors-1];
493 //Create processor worker threads.
494 for( int i=1; i<processors; i++ ){
495 // Allocate memory for thread data.
496 string extension = toString(i) + ".temp";
498 preClusterData* tempPreCluster = new preClusterData(fastafile, namefile, groupfile, countfile, (newFName+extension), (newNName+extension), newMFile, groups, m, lines[i].start, lines[i].end, diffs, topdown, i);
499 pDataArray.push_back(tempPreCluster);
500 processIDS.push_back(i);
502 //MySeqSumThreadFunction is in header. It must be global or static to work with the threads.
503 //default security attributes, thread function name, argument to thread function, use default creation flags, returns the thread identifier
504 hThreadArray[i-1] = CreateThread(NULL, 0, MyPreclusterThreadFunction, pDataArray[i-1], 0, &dwThreadIdArray[i-1]);
508 //using the main process as a worker saves time and memory
509 num = driverGroups(newFName, newNName, newMFile, lines[0].start, lines[0].end, groups);
511 //Wait until all threads have terminated.
512 WaitForMultipleObjects(processors-1, hThreadArray, TRUE, INFINITE);
514 //Close all thread handles and free memory allocations.
515 for(int i=0; i < pDataArray.size(); i++){
516 if (pDataArray[i]->count != (pDataArray[i]->end-pDataArray[i]->start)) {
517 m->mothurOut("[ERROR]: process " + toString(i) + " only processed " + toString(pDataArray[i]->count) + " of " + toString(pDataArray[i]->end-pDataArray[i]->start) + " groups assigned to it, quitting. \n"); m->control_pressed = true;
519 for (int j = 0; j < pDataArray[i]->mapFileNames.size(); j++) {
520 outputNames.push_back(pDataArray[i]->mapFileNames[j]); outputTypes["map"].push_back(pDataArray[i]->mapFileNames[j]);
522 CloseHandle(hThreadArray[i]);
523 delete pDataArray[i];
528 //append output files
529 for(int i=0;i<processIDS.size();i++){
530 //newFName = m->getFullPathName(".\\" + newFName);
531 //newNName = m->getFullPathName(".\\" + newNName);
533 m->appendFiles((newFName + toString(processIDS[i]) + ".temp"), newFName);
534 m->mothurRemove((newFName + toString(processIDS[i]) + ".temp"));
536 m->appendFiles((newNName + toString(processIDS[i]) + ".temp"), newNName);
537 m->mothurRemove((newNName + toString(processIDS[i]) + ".temp"));
543 catch(exception& e) {
544 m->errorOut(e, "PreClusterCommand", "createProcessesGroups");
548 /**************************************************************************************************/
549 int PreClusterCommand::driverGroups(string newFFile, string newNFile, string newMFile, int start, int end, vector<string> groups){
554 //precluster each group
555 for (int i = start; i < end; i++) {
559 if (m->control_pressed) { return 0; }
561 m->mothurOutEndLine(); m->mothurOut("Processing group " + groups[i] + ":"); m->mothurOutEndLine();
563 map<string, string> thisNameMap;
564 vector<Sequence> thisSeqs;
565 if (groupfile != "") {
566 thisSeqs = parser->getSeqs(groups[i]);
567 }else if (countfile != "") {
568 thisSeqs = cparser->getSeqs(groups[i]);
570 if (namefile != "") { thisNameMap = parser->getNameMap(groups[i]); }
572 //fill alignSeqs with this groups info.
573 numSeqs = loadSeqs(thisNameMap, thisSeqs, groups[i]);
575 if (m->control_pressed) { return 0; }
577 if (diffs > length) { m->mothurOut("Error: diffs is greater than your sequence length."); m->mothurOutEndLine(); m->control_pressed = true; return 0; }
579 int count= process(newMFile+groups[i]+".map");
580 outputNames.push_back(newMFile+groups[i]+".map"); outputTypes["map"].push_back(newMFile+groups[i]+".map");
582 if (m->control_pressed) { return 0; }
584 m->mothurOut("Total number of sequences before pre.cluster was " + toString(alignSeqs.size()) + "."); m->mothurOutEndLine();
585 m->mothurOut("pre.cluster removed " + toString(count) + " sequences."); m->mothurOutEndLine(); m->mothurOutEndLine();
586 printData(newFFile, newNFile, groups[i]);
588 m->mothurOut("It took " + toString(time(NULL) - start) + " secs to cluster " + toString(numSeqs) + " sequences."); m->mothurOutEndLine();
594 catch(exception& e) {
595 m->errorOut(e, "PreClusterCommand", "driverGroups");
599 /**************************************************************************************************/
600 int PreClusterCommand::process(string newMapFile){
603 m->openOutputFile(newMapFile, out);
605 //sort seqs by number of identical seqs
606 if (topdown) { sort(alignSeqs.begin(), alignSeqs.end(), comparePriorityTopDown); }
607 else { sort(alignSeqs.begin(), alignSeqs.end(), comparePriorityDownTop); }
610 int numSeqs = alignSeqs.size();
613 //think about running through twice...
614 for (int i = 0; i < numSeqs; i++) {
616 if (alignSeqs[i].active) { //this sequence has not been merged yet
618 string chunk = alignSeqs[i].seq.getName() + "\t" + toString(alignSeqs[i].numIdentical) + "\t" + toString(0) + "\t" + alignSeqs[i].seq.getAligned() + "\n";
620 //try to merge it with all smaller seqs
621 for (int j = i+1; j < numSeqs; j++) {
623 if (m->control_pressed) { out.close(); return 0; }
625 if (alignSeqs[j].active) { //this sequence has not been merged yet
626 //are you within "diff" bases
627 int mismatch = calcMisMatches(alignSeqs[i].seq.getAligned(), alignSeqs[j].seq.getAligned());
629 if (mismatch <= diffs) {
631 alignSeqs[i].names += ',' + alignSeqs[j].names;
632 alignSeqs[i].numIdentical += alignSeqs[j].numIdentical;
634 chunk += alignSeqs[j].seq.getName() + "\t" + toString(alignSeqs[j].numIdentical) + "\t" + toString(mismatch) + "\t" + alignSeqs[j].seq.getAligned() + "\n";
636 alignSeqs[j].active = 0;
637 alignSeqs[j].numIdentical = 0;
643 //remove from active list
644 alignSeqs[i].active = 0;
646 out << "ideal_seq_" << (i+1) << '\t' << alignSeqs[i].numIdentical << endl << chunk << endl;;
649 if(i % 100 == 0) { m->mothurOutJustToScreen(toString(i) + "\t" + toString(numSeqs - count) + "\t" + toString(count)+"\n"); }
652 map<int, string> mapFile;
653 map<int, int> originalCount;
654 map<int, int>::iterator itCount;
655 for (int i = 0; i < numSeqs; i++) { mapFile[i] = ""; originalCount[i] = alignSeqs[i].numIdentical; }
657 //think about running through twice...
658 for (int i = 0; i < numSeqs; i++) {
660 //try to merge it into larger seqs
661 for (int j = i+1; j < numSeqs; j++) {
663 if (m->control_pressed) { out.close(); return 0; }
665 if (originalCount[j] > originalCount[i]) { //this sequence is more abundant than I am
666 //are you within "diff" bases
667 int mismatch = calcMisMatches(alignSeqs[i].seq.getAligned(), alignSeqs[j].seq.getAligned());
669 if (mismatch <= diffs) {
671 alignSeqs[j].names += ',' + alignSeqs[i].names;
672 alignSeqs[j].numIdentical += alignSeqs[i].numIdentical;
674 mapFile[j] = alignSeqs[i].seq.getName() + "\t" + toString(alignSeqs[i].numIdentical) + "\t" + toString(mismatch) + "\t" + alignSeqs[i].seq.getAligned() + "\n" + mapFile[i];
675 alignSeqs[i].numIdentical = 0;
676 originalCount.erase(i);
679 j+=numSeqs; //exit search, we merged this one in.
681 }//end abundance check
684 if(i % 100 == 0) { m->mothurOutJustToScreen(toString(i) + "\t" + toString(numSeqs - count) + "\t" + toString(count)+"\n"); }
687 for (int i = 0; i < numSeqs; i++) {
688 if (alignSeqs[i].numIdentical != 0) {
689 out << "ideal_seq_" << (i+1) << '\t' << alignSeqs[i].numIdentical << endl << alignSeqs[i].seq.getName() + "\t" + toString(alignSeqs[i].numIdentical) + "\t" + toString(0) + "\t" + alignSeqs[i].seq.getAligned() + "\n" << mapFile[i] << endl;
696 if(numSeqs % 100 != 0) { m->mothurOut(toString(numSeqs) + "\t" + toString(numSeqs - count) + "\t" + toString(count)); m->mothurOutEndLine(); }
701 catch(exception& e) {
702 m->errorOut(e, "PreClusterCommand", "process");
706 /**************************************************************************************************/
707 int PreClusterCommand::readFASTA(){
712 m->openInputFile(fastafile, inFasta);
715 while (!inFasta.eof()) {
717 if (m->control_pressed) { inFasta.close(); return 0; }
719 Sequence seq(inFasta); m->gobble(inFasta);
721 if (seq.getName() != "") { //can get "" if commented line is at end of fasta file
722 if (namefile != "") {
723 itSize = sizes.find(seq.getName());
725 if (itSize == sizes.end()) { m->mothurOut(seq.getName() + " is not in your names file, please correct."); m->mothurOutEndLine(); exit(1); }
727 seqPNode tempNode(itSize->second, seq, names[seq.getName()]);
728 alignSeqs.push_back(tempNode);
729 lengths.insert(seq.getAligned().length());
731 }else { //no names file, you are identical to yourself
733 if (countfile != "") { numRep = ct.getNumSeqs(seq.getName()); }
734 seqPNode tempNode(numRep, seq, seq.getName());
735 alignSeqs.push_back(tempNode);
736 lengths.insert(seq.getAligned().length());
742 if (lengths.size() > 1) { m->control_pressed = true; m->mothurOut("[ERROR]: your sequences are not all the same length. pre.cluster requires sequences to be aligned."); m->mothurOutEndLine(); }
743 else if (lengths.size() == 1) { length = *(lengths.begin()); }
745 return alignSeqs.size();
748 catch(exception& e) {
749 m->errorOut(e, "PreClusterCommand", "readFASTA");
753 /**************************************************************************************************/
754 int PreClusterCommand::loadSeqs(map<string, string>& thisName, vector<Sequence>& thisSeqs, string group){
758 map<string, string>::iterator it;
760 map<string, int> thisCount;
761 if (countfile != "") { thisCount = cparser->getCountTable(group); }
763 for (int i = 0; i < thisSeqs.size(); i++) {
765 if (m->control_pressed) { return 0; }
767 if (namefile != "") {
768 it = thisName.find(thisSeqs[i].getName());
770 //should never be true since parser checks for this
771 if (it == thisName.end()) { m->mothurOut(thisSeqs[i].getName() + " is not in your names file, please correct."); m->mothurOutEndLine(); error = true; }
775 for(int j=0;j<(it->second).length();j++){
776 if((it->second)[j] == ','){ numReps++; }
779 seqPNode tempNode(numReps, thisSeqs[i], it->second);
780 alignSeqs.push_back(tempNode);
781 lengths.insert(thisSeqs[i].getAligned().length());
783 }else { //no names file, you are identical to yourself
785 if (countfile != "") {
786 map<string, int>::iterator it2 = thisCount.find(thisSeqs[i].getName());
788 //should never be true since parser checks for this
789 if (it2 == thisCount.end()) { m->mothurOut(thisSeqs[i].getName() + " is not in your count file, please correct."); m->mothurOutEndLine(); error = true; }
790 else { numRep = it2->second; }
792 seqPNode tempNode(numRep, thisSeqs[i], thisSeqs[i].getName());
793 alignSeqs.push_back(tempNode);
794 lengths.insert(thisSeqs[i].getAligned().length());
798 if (lengths.size() > 1) { error = true; m->mothurOut("[ERROR]: your sequences are not all the same length. pre.cluster requires sequences to be aligned."); m->mothurOutEndLine(); }
799 else if (lengths.size() == 1) { length = *(lengths.begin()); }
802 if (error) { m->control_pressed = true; }
806 return alignSeqs.size();
809 catch(exception& e) {
810 m->errorOut(e, "PreClusterCommand", "loadSeqs");
815 /**************************************************************************************************/
817 int PreClusterCommand::calcMisMatches(string seq1, string seq2){
821 for (int i = 0; i < seq1.length(); i++) {
823 if (seq1[i] != seq2[i]) { numBad++; }
824 if (numBad > diffs) { return length; } //to far to cluster
829 catch(exception& e) {
830 m->errorOut(e, "PreClusterCommand", "calcMisMatches");
834 /**************************************************************************************************/
836 int PreClusterCommand::mergeGroupCounts(string newcount, string newname, string newfasta){
839 m->openInputFile(newname, inNames);
841 string group, first, second;
842 set<string> uniqueNames;
843 while (!inNames.eof()) {
844 if (m->control_pressed) { break; }
845 inNames >> group; m->gobble(inNames);
846 inNames >> first; m->gobble(inNames);
847 inNames >> second; m->gobble(inNames);
849 vector<string> names;
850 m->splitAtComma(second, names);
852 uniqueNames.insert(first);
854 int total = ct.getGroupCount(first, group);
855 for (int i = 1; i < names.size(); i++) {
856 total += ct.getGroupCount(names[i], group);
857 ct.setAbund(names[i], group, 0);
859 ct.setAbund(first, group, total);
863 vector<string> namesOfSeqs = ct.getNamesOfSeqs();
864 for (int i = 0; i < namesOfSeqs.size(); i++) {
865 if (ct.getNumSeqs(namesOfSeqs[i]) == 0) {
866 ct.remove(namesOfSeqs[i]);
870 ct.printTable(newcount);
871 m->mothurRemove(newname);
873 if (bygroup) { //if by group, must remove the duplicate seqs that are named the same
875 m->openInputFile(newfasta, in);
878 m->openOutputFile(newfasta+"temp", out);
883 if (m->control_pressed) { break; }
885 Sequence seq(in); m->gobble(in);
887 if (seq.getName() != "") {
889 if (already.count(seq.getName()) == 0) {
890 seq.printSequence(out);
891 already.insert(seq.getName());
897 m->mothurRemove(newfasta);
898 m->renameFile(newfasta+"temp", newfasta);
903 catch(exception& e) {
904 m->errorOut(e, "PreClusterCommand", "mergeGroupCounts");
909 /**************************************************************************************************/
911 void PreClusterCommand::printData(string newfasta, string newname, string group){
917 m->openOutputFileAppend(newfasta, outFasta);
918 m->openOutputFileAppend(newname, outNames);
920 m->openOutputFile(newfasta, outFasta);
921 m->openOutputFile(newname, outNames);
924 if ((countfile != "") && (group == "")) { outNames << "Representative_Sequence\ttotal\n"; }
925 for (int i = 0; i < alignSeqs.size(); i++) {
926 if (alignSeqs[i].numIdentical != 0) {
927 alignSeqs[i].seq.printSequence(outFasta);
928 if (countfile != "") {
929 if (group != "") { outNames << group << '\t' << alignSeqs[i].seq.getName() << '\t' << alignSeqs[i].names << endl; }
930 else { outNames << alignSeqs[i].seq.getName() << '\t' << alignSeqs[i].numIdentical << endl; }
931 }else { outNames << alignSeqs[i].seq.getName() << '\t' << alignSeqs[i].names << endl; }
939 catch(exception& e) {
940 m->errorOut(e, "PreClusterCommand", "printData");
944 /**************************************************************************************************/
946 void PreClusterCommand::readNameFile(){
949 m->openInputFile(namefile, in);
950 string firstCol, secondCol;
953 in >> firstCol >> secondCol; m->gobble(in);
955 m->checkName(firstCol);
956 m->checkName(secondCol);
957 int size = m->getNumNames(secondCol);
959 names[firstCol] = secondCol;
960 sizes[firstCol] = size;
964 catch(exception& e) {
965 m->errorOut(e, "PreClusterCommand", "readNameFile");
970 /**************************************************************************************************/