2 * chimeraccodecommand.cpp
5 * Created by westcott on 3/30/10.
6 * Copyright 2010 Schloss Lab. All rights reserved.
10 #include "chimeraccodecommand.h"
13 //**********************************************************************************************************************
14 vector<string> ChimeraCcodeCommand::getValidParameters(){
16 string AlignArray[] = {"fasta", "filter", "processors", "window", "template", "mask", "numwanted", "outputdir","inputdir" };
17 vector<string> myArray (AlignArray, AlignArray+(sizeof(AlignArray)/sizeof(string)));
21 m->errorOut(e, "ChimeraCcodeCommand", "getValidParameters");
25 //**********************************************************************************************************************
26 ChimeraCcodeCommand::ChimeraCcodeCommand(){
28 abort = true; calledHelp = true;
29 vector<string> tempOutNames;
30 outputTypes["chimera"] = tempOutNames;
31 outputTypes["mapinfo"] = tempOutNames;
32 outputTypes["accnos"] = tempOutNames;
35 m->errorOut(e, "ChimeraCcodeCommand", "ChimeraCcodeCommand");
39 //**********************************************************************************************************************
40 vector<string> ChimeraCcodeCommand::getRequiredParameters(){
42 string AlignArray[] = {"template","fasta"};
43 vector<string> myArray (AlignArray, AlignArray+(sizeof(AlignArray)/sizeof(string)));
47 m->errorOut(e, "ChimeraCcodeCommand", "getRequiredParameters");
51 //**********************************************************************************************************************
52 vector<string> ChimeraCcodeCommand::getRequiredFiles(){
54 vector<string> myArray;
58 m->errorOut(e, "ChimeraCcodeCommand", "getRequiredFiles");
62 //***************************************************************************************************************
63 ChimeraCcodeCommand::ChimeraCcodeCommand(string option) {
65 abort = false; calledHelp = false;
67 //allow user to run help
68 if(option == "help") { help(); abort = true; calledHelp = true; }
71 //valid paramters for this command
72 string Array[] = {"fasta", "filter", "processors", "window", "template", "mask", "numwanted", "outputdir","inputdir" };
73 vector<string> myArray (Array, Array+(sizeof(Array)/sizeof(string)));
75 OptionParser parser(option);
76 map<string,string> parameters = parser.getParameters();
78 ValidParameters validParameter("chimera.ccode");
79 map<string,string>::iterator it;
81 //check to make sure all parameters are valid for command
82 for (it = parameters.begin(); it != parameters.end(); it++) {
83 if (validParameter.isValidParameter(it->first, myArray, it->second) != true) { abort = true; }
86 vector<string> tempOutNames;
87 outputTypes["chimera"] = tempOutNames;
88 outputTypes["mapinfo"] = tempOutNames;
89 outputTypes["accnos"] = tempOutNames;
91 //if the user changes the input directory command factory will send this info to us in the output parameter
92 string inputDir = validParameter.validFile(parameters, "inputdir", false);
93 if (inputDir == "not found"){ inputDir = ""; }
96 it = parameters.find("template");
97 //user has given a template file
98 if(it != parameters.end()){
99 path = m->hasPath(it->second);
100 //if the user has not given a path then, add inputdir. else leave path alone.
101 if (path == "") { parameters["template"] = inputDir + it->second; }
105 //check for required parameters
106 fastafile = validParameter.validFile(parameters, "fasta", false);
107 if (fastafile == "not found") { fastafile = ""; m->mothurOut("fasta is a required parameter for the chimera.ccode command."); m->mothurOutEndLine(); abort = true; }
109 m->splitAtDash(fastafile, fastaFileNames);
111 //go through files and make sure they are good, if not, then disregard them
112 for (int i = 0; i < fastaFileNames.size(); i++) {
113 if (inputDir != "") {
114 string path = m->hasPath(fastaFileNames[i]);
115 //if the user has not given a path then, add inputdir. else leave path alone.
116 if (path == "") { fastaFileNames[i] = inputDir + fastaFileNames[i]; }
122 ableToOpen = m->openInputFile(fastaFileNames[i], in, "noerror");
124 //if you can't open it, try default location
125 if (ableToOpen == 1) {
126 if (m->getDefaultPath() != "") { //default path is set
127 string tryPath = m->getDefaultPath() + m->getSimpleName(fastaFileNames[i]);
128 m->mothurOut("Unable to open " + fastaFileNames[i] + ". Trying default " + tryPath); m->mothurOutEndLine();
130 ableToOpen = m->openInputFile(tryPath, in2, "noerror");
132 fastaFileNames[i] = tryPath;
136 //if you can't open it, try default location
137 if (ableToOpen == 1) {
138 if (m->getOutputDir() != "") { //default path is set
139 string tryPath = m->getOutputDir() + m->getSimpleName(fastaFileNames[i]);
140 m->mothurOut("Unable to open " + fastaFileNames[i] + ". Trying output directory " + tryPath); m->mothurOutEndLine();
142 ableToOpen = m->openInputFile(tryPath, in2, "noerror");
144 fastaFileNames[i] = tryPath;
150 if (ableToOpen == 1) {
151 m->mothurOut("Unable to open " + fastaFileNames[i] + ". It will be disregarded."); m->mothurOutEndLine();
152 //erase from file list
153 fastaFileNames.erase(fastaFileNames.begin()+i);
158 //make sure there is at least one valid file left
159 if (fastaFileNames.size() == 0) { m->mothurOut("no valid files."); m->mothurOutEndLine(); abort = true; }
162 //if the user changes the output directory command factory will send this info to us in the output parameter
163 outputDir = validParameter.validFile(parameters, "outputdir", false); if (outputDir == "not found"){ outputDir = ""; }
165 templatefile = validParameter.validFile(parameters, "template", true);
166 if (templatefile == "not open") { abort = true; }
167 else if (templatefile == "not found") { templatefile = ""; m->mothurOut("template is a required parameter for the chimera.ccode command."); m->mothurOutEndLine(); abort = true; }
169 maskfile = validParameter.validFile(parameters, "mask", false);
170 if (maskfile == "not found") { maskfile = ""; }
171 else if (maskfile != "default") {
172 if (inputDir != "") {
173 string path = m->hasPath(maskfile);
174 //if the user has not given a path then, add inputdir. else leave path alone.
175 if (path == "") { maskfile = inputDir + maskfile; }
179 int ableToOpen = m->openInputFile(maskfile, in);
180 if (ableToOpen == 1) { abort = true; }
185 temp = validParameter.validFile(parameters, "filter", false); if (temp == "not found") { temp = "F"; }
186 filter = m->isTrue(temp);
188 temp = validParameter.validFile(parameters, "processors", false); if (temp == "not found") { temp = "1"; }
189 convert(temp, processors);
191 temp = validParameter.validFile(parameters, "window", false); if (temp == "not found") { temp = "0"; }
192 convert(temp, window);
194 temp = validParameter.validFile(parameters, "numwanted", false); if (temp == "not found") { temp = "20"; }
195 convert(temp, numwanted);
199 catch(exception& e) {
200 m->errorOut(e, "ChimeraCcodeCommand", "ChimeraCcodeCommand");
204 //**********************************************************************************************************************
206 void ChimeraCcodeCommand::help(){
209 m->mothurOut("The chimera.ccode command reads a fastafile and templatefile and outputs potentially chimeric sequences.\n");
210 m->mothurOut("This command was created using the algorythms described in the 'Evaluating putative chimeric sequences from PCR-amplified products' paper by Juan M. Gonzalez, Johannes Zimmerman and Cesareo Saiz-Jimenez.\n");
211 m->mothurOut("The chimera.ccode command parameters are fasta, template, filter, mask, processors, window and numwanted.\n");
212 m->mothurOut("The fasta parameter allows you to enter the fasta file containing your potentially chimeric sequences, and is required. \n");
213 m->mothurOut("You may enter multiple fasta files by separating their names with dashes. ie. fasta=abrecovery.fasta-amzon.fasta \n");
214 m->mothurOut("The template parameter allows you to enter a template file containing known non-chimeric sequences, and is required. \n");
215 m->mothurOut("The filter parameter allows you to specify if you would like to apply a vertical and 50% soft filter. \n");
216 m->mothurOut("The processors parameter allows you to specify how many processors you would like to use. The default is 1. \n");
218 m->mothurOut("When using MPI, the processors parameter is set to the number of MPI processes running. \n");
220 m->mothurOut("The mask parameter allows you to specify a file containing one sequence you wish to use as a mask for the your sequences. \n");
221 m->mothurOut("The window parameter allows you to specify the window size for searching for chimeras. \n");
222 m->mothurOut("The numwanted parameter allows you to specify how many sequences you would each query sequence compared with.\n");
223 m->mothurOut("The chimera.ccode command should be in the following format: \n");
224 m->mothurOut("chimera.ccode(fasta=yourFastaFile, template=yourTemplate) \n");
225 m->mothurOut("Example: chimera.ccode(fasta=AD.align, template=core_set_aligned.imputed.fasta) \n");
226 m->mothurOut("Note: No spaces between parameter labels (i.e. fasta), '=' and parameters (i.e.yourFastaFile).\n\n");
228 catch(exception& e) {
229 m->errorOut(e, "ChimeraCcodeCommand", "help");
234 //***************************************************************************************************************
236 ChimeraCcodeCommand::~ChimeraCcodeCommand(){ /* do nothing */ }
238 //***************************************************************************************************************
240 int ChimeraCcodeCommand::execute(){
243 if (abort == true) { if (calledHelp) { return 0; } return 2; }
245 for (int s = 0; s < fastaFileNames.size(); s++) {
247 m->mothurOut("Checking sequences from " + fastaFileNames[s] + " ..." ); m->mothurOutEndLine();
249 int start = time(NULL);
252 if (maskfile == "default") { m->mothurOut("I am using the default 236627 EU009184.1 Shigella dysenteriae str. FBD013."); m->mothurOutEndLine(); }
254 chimera = new Ccode(fastaFileNames[s], templatefile, filter, maskfile, window, numwanted, outputDir);
256 //is your template aligned?
257 if (chimera->getUnaligned()) { m->mothurOut("Your template sequences are different lengths, please correct."); m->mothurOutEndLine(); delete chimera; return 0; }
258 templateSeqsLength = chimera->getLength();
260 if (outputDir == "") { outputDir = m->hasPath(fastaFileNames[s]); }//if user entered a file with a path then preserve it
261 string outputFileName, accnosFileName;
262 if (maskfile != "") {
263 outputFileName = outputDir + m->getRootName(m->getSimpleName(fastaFileNames[s])) + maskfile + ".ccode.chimeras";
264 accnosFileName = outputDir + m->getRootName(m->getSimpleName(fastaFileNames[s])) + maskfile + ".ccode.accnos";
266 outputFileName = outputDir + m->getRootName(m->getSimpleName(fastaFileNames[s])) + "ccode.chimeras";
267 accnosFileName = outputDir + m->getRootName(m->getSimpleName(fastaFileNames[s])) + "ccode.accnos";
270 string mapInfo = outputDir + m->getRootName(m->getSimpleName(fastaFileNames[s])) + "mapinfo";
272 if (m->control_pressed) { delete chimera; for (int j = 0; j < outputNames.size(); j++) { remove(outputNames[j].c_str()); } outputTypes.clear(); return 0; }
276 int pid, numSeqsPerProcessor;
278 vector<unsigned long int> MPIPos;
281 MPI_Comm_rank(MPI_COMM_WORLD, &pid); //find out who we are
282 MPI_Comm_size(MPI_COMM_WORLD, &processors);
286 MPI_File outMPIAccnos;
288 int outMode=MPI_MODE_CREATE|MPI_MODE_WRONLY;
289 int inMode=MPI_MODE_RDONLY;
291 char outFilename[1024];
292 strcpy(outFilename, outputFileName.c_str());
294 char outAccnosFilename[1024];
295 strcpy(outAccnosFilename, accnosFileName.c_str());
297 char inFileName[1024];
298 strcpy(inFileName, fastaFileNames[s].c_str());
300 MPI_File_open(MPI_COMM_WORLD, inFileName, inMode, MPI_INFO_NULL, &inMPI); //comm, filename, mode, info, filepointer
301 MPI_File_open(MPI_COMM_WORLD, outFilename, outMode, MPI_INFO_NULL, &outMPI);
302 MPI_File_open(MPI_COMM_WORLD, outAccnosFilename, outMode, MPI_INFO_NULL, &outMPIAccnos);
304 if (m->control_pressed) { MPI_File_close(&inMPI); MPI_File_close(&outMPI); MPI_File_close(&outMPIAccnos); for (int j = 0; j < outputNames.size(); j++) { remove(outputNames[j].c_str()); } outputTypes.clear(); delete chimera; return 0; }
306 if (pid == 0) { //you are the root process
307 string outTemp = "For full window mapping info refer to " + mapInfo + "\n\n";
310 int length = outTemp.length();
311 char* buf2 = new char[length];
312 memcpy(buf2, outTemp.c_str(), length);
314 MPI_File_write_shared(outMPI, buf2, length, MPI_CHAR, &status);
317 MPIPos = m->setFilePosFasta(fastaFileNames[s], numSeqs); //fills MPIPos, returns numSeqs
319 //send file positions to all processes
320 for(int i = 1; i < processors; i++) {
321 MPI_Send(&numSeqs, 1, MPI_INT, i, tag, MPI_COMM_WORLD);
322 MPI_Send(&MPIPos[0], (numSeqs+1), MPI_LONG, i, tag, MPI_COMM_WORLD);
325 //figure out how many sequences you have to align
326 numSeqsPerProcessor = numSeqs / processors;
327 int startIndex = pid * numSeqsPerProcessor;
328 if(pid == (processors - 1)){ numSeqsPerProcessor = numSeqs - pid * numSeqsPerProcessor; }
332 driverMPI(startIndex, numSeqsPerProcessor, inMPI, outMPI, outMPIAccnos, MPIPos);
334 if (m->control_pressed) { MPI_File_close(&inMPI); MPI_File_close(&outMPI); MPI_File_close(&outMPIAccnos); remove(outputFileName.c_str()); remove(accnosFileName.c_str()); for (int j = 0; j < outputNames.size(); j++) { remove(outputNames[j].c_str()); } outputTypes.clear(); delete chimera; return 0; }
336 }else{ //you are a child process
337 MPI_Recv(&numSeqs, 1, MPI_INT, 0, tag, MPI_COMM_WORLD, &status);
338 MPIPos.resize(numSeqs+1);
339 MPI_Recv(&MPIPos[0], (numSeqs+1), MPI_LONG, 0, tag, MPI_COMM_WORLD, &status);
341 //figure out how many sequences you have to align
342 numSeqsPerProcessor = numSeqs / processors;
343 int startIndex = pid * numSeqsPerProcessor;
344 if(pid == (processors - 1)){ numSeqsPerProcessor = numSeqs - pid * numSeqsPerProcessor; }
348 driverMPI(startIndex, numSeqsPerProcessor, inMPI, outMPI, outMPIAccnos, MPIPos);
350 if (m->control_pressed) { MPI_File_close(&inMPI); MPI_File_close(&outMPI); MPI_File_close(&outMPIAccnos); for (int j = 0; j < outputNames.size(); j++) { remove(outputNames[j].c_str()); } outputTypes.clear(); delete chimera; return 0; }
354 MPI_File_close(&inMPI);
355 MPI_File_close(&outMPI);
356 MPI_File_close(&outMPIAccnos);
358 MPI_Barrier(MPI_COMM_WORLD); //make everyone wait - just in case
362 string tempHeader = outputDir + m->getRootName(m->getSimpleName(fastaFileNames[s])) + maskfile + "ccode.chimeras.tempHeader";
363 m->openOutputFile(tempHeader, outHeader);
365 outHeader << "For full window mapping info refer to " << mapInfo << endl << endl;
369 vector<unsigned long int> positions = m->divideFile(fastaFileNames[s], processors);
371 for (int i = 0; i < (positions.size()-1); i++) {
372 lines.push_back(new linePair(positions[i], positions[(i+1)]));
376 #if defined (__APPLE__) || (__MACH__) || (linux) || (__linux)
379 numSeqs = driver(lines[0], outputFileName, fastaFileNames[s], accnosFileName);
381 if (m->control_pressed) { remove(outputFileName.c_str()); remove(tempHeader.c_str()); remove(accnosFileName.c_str()); for (int j = 0; j < outputNames.size(); j++) { remove(outputNames[j].c_str()); } for (int i = 0; i < lines.size(); i++) { delete lines[i]; } outputTypes.clear(); lines.clear(); delete chimera; return 0; }
384 processIDS.resize(0);
386 numSeqs = createProcesses(outputFileName, fastaFileNames[s], accnosFileName);
388 rename((outputFileName + toString(processIDS[0]) + ".temp").c_str(), outputFileName.c_str());
389 rename((accnosFileName + toString(processIDS[0]) + ".temp").c_str(), accnosFileName.c_str());
391 //append output files
392 for(int i=1;i<processors;i++){
393 m->appendFiles((outputFileName + toString(processIDS[i]) + ".temp"), outputFileName);
394 remove((outputFileName + toString(processIDS[i]) + ".temp").c_str());
397 //append output files
398 for(int i=1;i<processors;i++){
399 m->appendFiles((accnosFileName + toString(processIDS[i]) + ".temp"), accnosFileName);
400 remove((accnosFileName + toString(processIDS[i]) + ".temp").c_str());
403 if (m->control_pressed) {
404 remove(outputFileName.c_str());
405 remove(accnosFileName.c_str());
406 for (int j = 0; j < outputNames.size(); j++) { remove(outputNames[j].c_str()); } outputTypes.clear();
407 for (int i = 0; i < lines.size(); i++) { delete lines[i]; } lines.clear();
415 numSeqs = driver(lines[0], outputFileName, fastaFileNames[s], accnosFileName);
417 if (m->control_pressed) { remove(outputFileName.c_str()); remove(tempHeader.c_str()); remove(accnosFileName.c_str()); for (int j = 0; j < outputNames.size(); j++) { remove(outputNames[j].c_str()); } for (int i = 0; i < lines.size(); i++) { delete lines[i]; } outputTypes.clear(); lines.clear(); delete chimera; return 0; }
421 m->appendFiles(outputFileName, tempHeader);
423 remove(outputFileName.c_str());
424 rename(tempHeader.c_str(), outputFileName.c_str());
429 outputNames.push_back(outputFileName); outputTypes["chimera"].push_back(outputFileName);
430 outputNames.push_back(mapInfo); outputTypes["mapinfo"].push_back(mapInfo);
431 outputNames.push_back(accnosFileName); outputTypes["accnos"].push_back(accnosFileName);
433 for (int i = 0; i < lines.size(); i++) { delete lines[i]; } lines.clear();
435 m->mothurOutEndLine(); m->mothurOut("It took " + toString(time(NULL) - start) + " secs to check " + toString(numSeqs) + " sequences."); m->mothurOutEndLine();
438 m->mothurOutEndLine();
439 m->mothurOut("Output File Names: "); m->mothurOutEndLine();
440 for (int i = 0; i < outputNames.size(); i++) { m->mothurOut(outputNames[i]); m->mothurOutEndLine(); }
441 m->mothurOutEndLine();
446 catch(exception& e) {
447 m->errorOut(e, "ChimeraCcodeCommand", "execute");
451 //**********************************************************************************************************************
453 int ChimeraCcodeCommand::driver(linePair* filePos, string outputFName, string filename, string accnos){
456 m->openOutputFile(outputFName, out);
459 m->openOutputFile(accnos, out2);
462 m->openInputFile(filename, inFASTA);
464 inFASTA.seekg(filePos->start);
471 if (m->control_pressed) { return 1; }
473 Sequence* candidateSeq = new Sequence(inFASTA); m->gobble(inFASTA);
475 if (candidateSeq->getName() != "") { //incase there is a commented sequence at the end of a file
477 if (candidateSeq->getAligned().length() != templateSeqsLength) {
478 m->mothurOut(candidateSeq->getName() + " is not the same length as the template sequences. Skipping."); m->mothurOutEndLine();
481 chimera->getChimeras(candidateSeq);
483 if (m->control_pressed) { delete candidateSeq; return 1; }
486 chimera->print(out, out2);
492 #if defined (__APPLE__) || (__MACH__) || (linux) || (__linux)
493 unsigned long int pos = inFASTA.tellg();
494 if ((pos == -1) || (pos >= filePos->end)) { break; }
496 if (inFASTA.eof()) { break; }
500 if((count) % 100 == 0){ m->mothurOut("Processing sequence: " + toString(count)); m->mothurOutEndLine(); }
503 if((count) % 100 != 0){ m->mothurOut("Processing sequence: " + toString(count)); m->mothurOutEndLine(); }
511 catch(exception& e) {
512 m->errorOut(e, "ChimeraCcodeCommand", "driver");
516 //**********************************************************************************************************************
518 int ChimeraCcodeCommand::driverMPI(int start, int num, MPI_File& inMPI, MPI_File& outMPI, MPI_File& outAccMPI, vector<unsigned long int>& MPIPos){
523 MPI_Comm_rank(MPI_COMM_WORLD, &pid); //find out who we are
525 for(int i=0;i<num;i++){
527 if (m->control_pressed) { return 0; }
530 int length = MPIPos[start+i+1] - MPIPos[start+i];
532 char* buf4 = new char[length];
534 MPI_File_read_at(inMPI, MPIPos[start+i], buf4, length, MPI_CHAR, &status);
536 string tempBuf = buf4;
537 if (tempBuf.length() > length) { tempBuf = tempBuf.substr(0, length); }
538 istringstream iss (tempBuf,istringstream::in);
541 Sequence* candidateSeq = new Sequence(iss); m->gobble(iss);
543 if (candidateSeq->getName() != "") { //incase there is a commented sequence at the end of a file
545 if (candidateSeq->getAligned().length() != templateSeqsLength) {
546 m->mothurOut(candidateSeq->getName() + " is not the same length as the template sequences. Skipping."); m->mothurOutEndLine();
549 chimera->getChimeras(candidateSeq);
551 if (m->control_pressed) { delete candidateSeq; return 1; }
554 chimera->print(outMPI, outAccMPI);
560 if((i+1) % 100 == 0){ cout << "Processing sequence: " << (i+1) << endl; m->mothurOutJustToLog("Processing sequence: " + toString(i+1) + "\n"); }
563 if(num % 100 != 0){ cout << "Processing sequence: " << num << endl; m->mothurOutJustToLog("Processing sequence: " + toString(num) + "\n"); }
568 catch(exception& e) {
569 m->errorOut(e, "ChimeraCcodeCommand", "driverMPI");
575 /**************************************************************************************************/
577 int ChimeraCcodeCommand::createProcesses(string outputFileName, string filename, string accnos) {
579 #if defined (__APPLE__) || (__MACH__) || (linux) || (__linux)
583 //loop through and create all the processes you want
584 while (process != processors) {
588 processIDS.push_back(pid); //create map from line number to pid so you can append files in correct order later
591 num = driver(lines[process], outputFileName + toString(getpid()) + ".temp", filename, accnos + toString(getpid()) + ".temp");
593 //pass numSeqs to parent
595 string tempFile = outputFileName + toString(getpid()) + ".num.temp";
596 m->openOutputFile(tempFile, out);
602 m->mothurOut("[ERROR]: unable to spawn the necessary processes."); m->mothurOutEndLine();
603 for (int i = 0; i < processIDS.size(); i++) { kill (processIDS[i], SIGINT); }
608 //force parent to wait until all the processes are done
609 for (int i=0;i<processors;i++) {
610 int temp = processIDS[i];
614 for (int i = 0; i < processIDS.size(); i++) {
616 string tempFile = outputFileName + toString(processIDS[i]) + ".num.temp";
617 m->openInputFile(tempFile, in);
618 if (!in.eof()) { int tempNum = 0; in >> tempNum; num += tempNum; }
619 in.close(); remove(tempFile.c_str());
625 catch(exception& e) {
626 m->errorOut(e, "ChimeraCcodeCommand", "createProcesses");
630 //**********************************************************************************************************************