2 * chimerapintailcommand.cpp
5 * Created by westcott on 4/1/10.
6 * Copyright 2010 Schloss Lab. All rights reserved.
10 #include "chimerapintailcommand.h"
13 //***************************************************************************************************************
15 ChimeraPintailCommand::ChimeraPintailCommand(string option) {
19 //allow user to run help
20 if(option == "help") { help(); abort = true; }
23 //valid paramters for this command
24 string Array[] = {"fasta","filter","processors","window" ,"increment","template","conservation","quantile","mask","outputdir","inputdir"};
25 vector<string> myArray (Array, Array+(sizeof(Array)/sizeof(string)));
27 OptionParser parser(option);
28 map<string,string> parameters = parser.getParameters();
30 ValidParameters validParameter;
31 map<string,string>::iterator it;
33 //check to make sure all parameters are valid for command
34 for (it = parameters.begin(); it != parameters.end(); it++) {
35 if (validParameter.isValidParameter(it->first, myArray, it->second) != true) { abort = true; }
38 //if the user changes the input directory command factory will send this info to us in the output parameter
39 string inputDir = validParameter.validFile(parameters, "inputdir", false);
40 if (inputDir == "not found"){ inputDir = ""; }
43 it = parameters.find("fasta");
44 //user has given a template file
45 if(it != parameters.end()){
46 path = hasPath(it->second);
47 //if the user has not given a path then, add inputdir. else leave path alone.
48 if (path == "") { parameters["fasta"] = inputDir + it->second; }
51 it = parameters.find("template");
52 //user has given a template file
53 if(it != parameters.end()){
54 path = hasPath(it->second);
55 //if the user has not given a path then, add inputdir. else leave path alone.
56 if (path == "") { parameters["template"] = inputDir + it->second; }
59 it = parameters.find("conservation");
60 //user has given a template file
61 if(it != parameters.end()){
62 path = hasPath(it->second);
63 //if the user has not given a path then, add inputdir. else leave path alone.
64 if (path == "") { parameters["conservation"] = inputDir + it->second; }
67 it = parameters.find("quantile");
68 //user has given a template file
69 if(it != parameters.end()){
70 path = hasPath(it->second);
71 //if the user has not given a path then, add inputdir. else leave path alone.
72 if (path == "") { parameters["quantile"] = inputDir + it->second; }
77 //check for required parameters
78 fastafile = validParameter.validFile(parameters, "fasta", true);
79 if (fastafile == "not open") { abort = true; }
80 else if (fastafile == "not found") { fastafile = ""; m->mothurOut("fasta is a required parameter for the chimera.pintail command."); m->mothurOutEndLine(); abort = true; }
82 //if the user changes the output directory command factory will send this info to us in the output parameter
83 outputDir = validParameter.validFile(parameters, "outputdir", false); if (outputDir == "not found"){
85 outputDir += hasPath(fastafile); //if user entered a file with a path then preserve it
88 templatefile = validParameter.validFile(parameters, "template", true);
89 if (templatefile == "not open") { abort = true; }
90 else if (templatefile == "not found") { templatefile = ""; m->mothurOut("template is a required parameter for the chimera.pintail command."); m->mothurOutEndLine(); abort = true; }
92 consfile = validParameter.validFile(parameters, "conservation", true);
93 if (consfile == "not open") { abort = true; }
94 else if (consfile == "not found") { consfile = ""; }
96 quanfile = validParameter.validFile(parameters, "quantile", true);
97 if (quanfile == "not open") { abort = true; }
98 else if (quanfile == "not found") { quanfile = ""; }
100 maskfile = validParameter.validFile(parameters, "mask", false);
101 if (maskfile == "not found") { maskfile = ""; }
102 else if (maskfile != "default") {
103 if (inputDir != "") {
104 string path = hasPath(maskfile);
105 //if the user has not given a path then, add inputdir. else leave path alone.
106 if (path == "") { maskfile = inputDir + maskfile; }
110 int ableToOpen = openInputFile(maskfile, in);
111 if (ableToOpen == 1) { abort = true; }
116 temp = validParameter.validFile(parameters, "filter", false); if (temp == "not found") { temp = "F"; }
117 filter = isTrue(temp);
119 temp = validParameter.validFile(parameters, "processors", false); if (temp == "not found") { temp = "1"; }
120 convert(temp, processors);
122 temp = validParameter.validFile(parameters, "window", false); if (temp == "not found") { temp = "0"; }
123 convert(temp, window);
125 temp = validParameter.validFile(parameters, "increment", false); if (temp == "not found") { temp = "25"; }
126 convert(temp, increment);
129 catch(exception& e) {
130 m->errorOut(e, "ChimeraPintailCommand", "ChimeraPintailCommand");
134 //**********************************************************************************************************************
136 void ChimeraPintailCommand::help(){
139 m->mothurOut("The chimera.pintail command reads a fastafile and templatefile and outputs potentially chimeric sequences.\n");
140 m->mothurOut("This command was created using the algorythms described in the 'At Least 1 in 20 16S rRNA Sequence Records Currently Held in the Public Repositories is Estimated To Contain Substantial Anomalies' paper by Kevin E. Ashelford 1, Nadia A. Chuzhanova 3, John C. Fry 1, Antonia J. Jones 2 and Andrew J. Weightman 1.\n");
141 m->mothurOut("The chimera.pintail command parameters are fasta, template, filter, mask, processors, window, increment, conservation and quantile.\n");
142 m->mothurOut("The fasta parameter allows you to enter the fasta file containing your potentially chimeric sequences, and is required. \n");
143 m->mothurOut("The template parameter allows you to enter a template file containing known non-chimeric sequences, and is required. \n");
144 m->mothurOut("The filter parameter allows you to specify if you would like to apply a vertical and 50% soft filter. \n");
145 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, by default no mask is applied. You can apply an ecoli mask by typing, mask=default. \n");
146 m->mothurOut("The processors parameter allows you to specify how many processors you would like to use. The default is 1. \n");
148 m->mothurOut("When using MPI, the processors parameter is set to the number of MPI processes running. \n");
150 m->mothurOut("The window parameter allows you to specify the window size for searching for chimeras, default=300. \n");
151 m->mothurOut("The increment parameter allows you to specify how far you move each window while finding chimeric sequences, default=25.\n");
152 m->mothurOut("The conservation parameter allows you to enter a frequency file containing the highest bases frequency at each place in the alignment.\n");
153 m->mothurOut("The quantile parameter allows you to enter a file containing quantiles for a template files sequences, if you use the filter the quantile file generated becomes unique to the fasta file you used.\n");
154 m->mothurOut("The chimera.pintail command should be in the following format: \n");
155 m->mothurOut("chimera.seqs(fasta=yourFastaFile, filter=yourFilter, correction=yourCorrection, processors=yourProcessors, method=bellerophon) \n");
156 m->mothurOut("Example: chimera.seqs(fasta=AD.align, filter=True, correction=true, method=bellerophon, window=200) \n");
157 m->mothurOut("Note: No spaces between parameter labels (i.e. fasta), '=' and parameters (i.e.yourFastaFile).\n\n");
159 catch(exception& e) {
160 m->errorOut(e, "ChimeraPintailCommand", "help");
165 //***************************************************************************************************************
167 ChimeraPintailCommand::~ChimeraPintailCommand(){ /* do nothing */ }
169 //***************************************************************************************************************
171 int ChimeraPintailCommand::execute(){
174 if (abort == true) { return 0; }
176 int start = time(NULL);
179 if (maskfile == "default") { m->mothurOut("I am using the default 236627 EU009184.1 Shigella dysenteriae str. FBD013."); m->mothurOutEndLine(); }
181 chimera = new Pintail(fastafile, templatefile, filter, processors, maskfile, consfile, quanfile, window, increment, outputDir);
183 string outputFileName, accnosFileName;
184 if (maskfile != "") {
185 outputFileName = outputDir + getRootName(getSimpleName(fastafile)) + maskfile + ".pintail.chimeras";
186 accnosFileName = outputDir + getRootName(getSimpleName(fastafile)) + maskfile + ".pintail.accnos";
188 outputFileName = outputDir + getRootName(getSimpleName(fastafile)) + "pintail.chimeras";
189 accnosFileName = outputDir + getRootName(getSimpleName(fastafile)) + "pintail.accnos";
191 bool hasAccnos = true;
193 if (m->control_pressed) { delete chimera; return 0; }
195 if (chimera->getUnaligned()) {
196 m->mothurOut("Your template sequences are different lengths, please correct."); m->mothurOutEndLine();
200 templateSeqsLength = chimera->getLength();
203 int pid, end, numSeqsPerProcessor;
206 MPIWroteAccnos = false;
209 MPI_Comm_rank(MPI_COMM_WORLD, &pid); //find out who we are
210 MPI_Comm_size(MPI_COMM_WORLD, &processors);
214 MPI_File outMPIAccnos;
216 int outMode=MPI_MODE_CREATE|MPI_MODE_WRONLY;
217 int inMode=MPI_MODE_RDONLY;
219 //char* outFilename = new char[outputFileName.length()];
220 //memcpy(outFilename, outputFileName.c_str(), outputFileName.length());
222 char outFilename[1024];
223 strcpy(outFilename, outputFileName.c_str());
225 //char* outAccnosFilename = new char[accnosFileName.length()];
226 //memcpy(outAccnosFilename, accnosFileName.c_str(), accnosFileName.length());
228 char outAccnosFilename[1024];
229 strcpy(outAccnosFilename, accnosFileName.c_str());
231 //char* inFileName = new char[fastafile.length()];
232 //memcpy(inFileName, fastafile.c_str(), fastafile.length());
234 char inFileName[1024];
235 strcpy(inFileName, fastafile.c_str());
237 MPI_File_open(MPI_COMM_WORLD, inFileName, inMode, MPI_INFO_NULL, &inMPI); //comm, filename, mode, info, filepointer
238 MPI_File_open(MPI_COMM_WORLD, outFilename, outMode, MPI_INFO_NULL, &outMPI);
239 MPI_File_open(MPI_COMM_WORLD, outAccnosFilename, outMode, MPI_INFO_NULL, &outMPIAccnos);
242 //delete outFilename;
243 //delete outAccnosFilename;
245 if (m->control_pressed) { MPI_File_close(&inMPI); MPI_File_close(&outMPI); MPI_File_close(&outMPIAccnos); delete chimera; return 0; }
247 if (pid == 0) { //you are the root process
249 MPIPos = setFilePosFasta(fastafile, numSeqs); //fills MPIPos, returns numSeqs
251 //send file positions to all processes
252 MPI_Bcast(&numSeqs, 1, MPI_INT, 0, MPI_COMM_WORLD); //send numSeqs
253 MPI_Bcast(&MPIPos[0], (numSeqs+1), MPI_LONG, 0, MPI_COMM_WORLD); //send file pos
255 //figure out how many sequences you have to align
256 numSeqsPerProcessor = numSeqs / processors;
257 int startIndex = pid * numSeqsPerProcessor;
258 if(pid == (processors - 1)){ numSeqsPerProcessor = numSeqs - pid * numSeqsPerProcessor; }
262 driverMPI(startIndex, numSeqsPerProcessor, inMPI, outMPI, outMPIAccnos, MPIPos);
264 if (m->control_pressed) { MPI_File_close(&inMPI); MPI_File_close(&outMPI); MPI_File_close(&outMPIAccnos); remove(outputFileName.c_str()); remove(accnosFileName.c_str()); delete chimera; return 0; }
266 for (int i = 1; i < processors; i++) {
268 MPI_Recv(&tempResult, 1, MPI_INT, i, tag, MPI_COMM_WORLD, &status);
269 if (tempResult != 0) { MPIWroteAccnos = true; }
271 }else{ //you are a child process
272 MPI_Bcast(&numSeqs, 1, MPI_INT, 0, MPI_COMM_WORLD); //get numSeqs
273 MPIPos.resize(numSeqs+1);
274 MPI_Bcast(&MPIPos[0], (numSeqs+1), MPI_LONG, 0, MPI_COMM_WORLD); //get file positions
276 //figure out how many sequences you have to align
277 numSeqsPerProcessor = numSeqs / processors;
278 int startIndex = pid * numSeqsPerProcessor;
279 if(pid == (processors - 1)){ numSeqsPerProcessor = numSeqs - pid * numSeqsPerProcessor; }
283 driverMPI(startIndex, numSeqsPerProcessor, inMPI, outMPI, outMPIAccnos, MPIPos);
285 if (m->control_pressed) { MPI_File_close(&inMPI); MPI_File_close(&outMPI); MPI_File_close(&outMPIAccnos); delete chimera; return 0; }
287 MPI_Send(&MPIWroteAccnos, 1, MPI_INT, 0, tag, MPI_COMM_WORLD);
291 MPI_File_close(&inMPI);
292 MPI_File_close(&outMPI);
293 MPI_File_close(&outMPIAccnos);
295 //delete accnos file if blank
297 if (!MPIWroteAccnos) {
299 //MPI_File_delete(outAccnosFilename, info);
301 remove(accnosFileName.c_str());
308 #if defined (__APPLE__) || (__MACH__) || (linux) || (__linux)
311 openInputFile(fastafile, inFASTA);
312 numSeqs=count(istreambuf_iterator<char>(inFASTA),istreambuf_iterator<char>(), '>');
315 lines.push_back(new linePair(0, numSeqs));
317 driver(lines[0], outputFileName, fastafile, accnosFileName);
319 if (m->control_pressed) {
320 remove(outputFileName.c_str());
321 remove(accnosFileName.c_str());
322 for (int i = 0; i < lines.size(); i++) { delete lines[i]; } lines.clear();
327 //delete accnos file if its blank
328 if (isBlank(accnosFileName)) { remove(accnosFileName.c_str()); hasAccnos = false; }
331 vector<int> positions;
332 processIDS.resize(0);
335 openInputFile(fastafile, inFASTA);
338 while(!inFASTA.eof()){
339 input = getline(inFASTA);
340 if (input.length() != 0) {
341 if(input[0] == '>'){ long int pos = inFASTA.tellg(); positions.push_back(pos - input.length() - 1); }
346 numSeqs = positions.size();
348 int numSeqsPerProcessor = numSeqs / processors;
350 for (int i = 0; i < processors; i++) {
351 long int startPos = positions[ i * numSeqsPerProcessor ];
352 if(i == processors - 1){
353 numSeqsPerProcessor = numSeqs - i * numSeqsPerProcessor;
355 lines.push_back(new linePair(startPos, numSeqsPerProcessor));
359 createProcesses(outputFileName, fastafile, accnosFileName);
361 rename((outputFileName + toString(processIDS[0]) + ".temp").c_str(), outputFileName.c_str());
363 //append output files
364 for(int i=1;i<processors;i++){
365 appendFiles((outputFileName + toString(processIDS[i]) + ".temp"), outputFileName);
366 remove((outputFileName + toString(processIDS[i]) + ".temp").c_str());
369 vector<string> nonBlankAccnosFiles;
370 //delete blank accnos files generated with multiple processes
371 for(int i=0;i<processors;i++){
372 if (!(isBlank(accnosFileName + toString(processIDS[i]) + ".temp"))) {
373 nonBlankAccnosFiles.push_back(accnosFileName + toString(processIDS[i]) + ".temp");
374 }else { remove((accnosFileName + toString(processIDS[i]) + ".temp").c_str()); }
377 //append accnos files
378 if (nonBlankAccnosFiles.size() != 0) {
379 rename(nonBlankAccnosFiles[0].c_str(), accnosFileName.c_str());
381 for (int h=1; h < nonBlankAccnosFiles.size(); h++) {
382 appendFiles(nonBlankAccnosFiles[h], accnosFileName);
383 remove(nonBlankAccnosFiles[h].c_str());
385 }else{ hasAccnos = false; }
387 if (m->control_pressed) {
388 remove(outputFileName.c_str());
389 remove(accnosFileName.c_str());
390 for (int i = 0; i < lines.size(); i++) { delete lines[i]; } lines.clear();
398 openInputFile(fastafile, inFASTA);
399 numSeqs=count(istreambuf_iterator<char>(inFASTA),istreambuf_iterator<char>(), '>');
401 lines.push_back(new linePair(0, numSeqs));
403 driver(lines[0], outputFileName, fastafile, accnosFileName);
405 if (m->control_pressed) {
406 remove(outputFileName.c_str());
407 remove(accnosFileName.c_str());
408 for (int i = 0; i < lines.size(); i++) { delete lines[i]; } lines.clear();
413 //delete accnos file if its blank
414 if (isBlank(accnosFileName)) { remove(accnosFileName.c_str()); hasAccnos = false; }
420 for (int i = 0; i < lines.size(); i++) { delete lines[i]; } lines.clear();
422 m->mothurOutEndLine();
423 m->mothurOut("Output File Names: "); m->mothurOutEndLine();
424 m->mothurOut(outputFileName); m->mothurOutEndLine();
425 if (hasAccnos) { m->mothurOut(accnosFileName); m->mothurOutEndLine(); }
426 m->mothurOutEndLine();
427 m->mothurOutEndLine(); m->mothurOut("It took " + toString(time(NULL) - start) + " secs to check " + toString(numSeqs) + " sequences."); m->mothurOutEndLine();
432 catch(exception& e) {
433 m->errorOut(e, "ChimeraPintailCommand", "execute");
437 //**********************************************************************************************************************
439 int ChimeraPintailCommand::driver(linePair* line, string outputFName, string filename, string accnos){
442 openOutputFile(outputFName, out);
445 openOutputFile(accnos, out2);
448 openInputFile(filename, inFASTA);
450 inFASTA.seekg(line->start);
452 for(int i=0;i<line->numSeqs;i++){
454 if (m->control_pressed) { return 1; }
456 Sequence* candidateSeq = new Sequence(inFASTA); gobble(inFASTA);
458 if (candidateSeq->getName() != "") { //incase there is a commented sequence at the end of a file
460 if (candidateSeq->getAligned().length() != templateSeqsLength) { //chimeracheck does not require seqs to be aligned
461 m->mothurOut(candidateSeq->getName() + " is not the same length as the template sequences. Skipping."); m->mothurOutEndLine();
464 chimera->getChimeras(candidateSeq);
466 if (m->control_pressed) { delete candidateSeq; return 1; }
469 chimera->print(out, out2);
475 if((i+1) % 100 == 0){ m->mothurOut("Processing sequence: " + toString(i+1)); m->mothurOutEndLine(); }
478 if((line->numSeqs) % 100 != 0){ m->mothurOut("Processing sequence: " + toString(line->numSeqs)); m->mothurOutEndLine(); }
486 catch(exception& e) {
487 m->errorOut(e, "ChimeraPintailCommand", "driver");
491 //**********************************************************************************************************************
493 int ChimeraPintailCommand::driverMPI(int start, int num, MPI_File& inMPI, MPI_File& outMPI, MPI_File& outAccMPI, vector<long>& MPIPos){
498 MPI_Comm_rank(MPI_COMM_WORLD, &pid); //find out who we are
500 for(int i=0;i<num;i++){
502 if (m->control_pressed) { return 1; }
505 int length = MPIPos[start+i+1] - MPIPos[start+i];
507 char* buf4 = new char[length];
508 MPI_File_read_at(inMPI, MPIPos[start+i], buf4, length, MPI_CHAR, &status);
510 string tempBuf = buf4;
511 if (tempBuf.length() > length) { tempBuf = tempBuf.substr(0, length); }
512 istringstream iss (tempBuf,istringstream::in);
515 Sequence* candidateSeq = new Sequence(iss); gobble(iss);
517 if (candidateSeq->getName() != "") { //incase there is a commented sequence at the end of a file
519 if (candidateSeq->getAligned().length() != templateSeqsLength) { //chimeracheck does not require seqs to be aligned
520 m->mothurOut(candidateSeq->getName() + " is not the same length as the template sequences. Skipping."); m->mothurOutEndLine();
523 chimera->getChimeras(candidateSeq);
525 if (m->control_pressed) { delete candidateSeq; return 1; }
528 bool isChimeric = chimera->print(outMPI, outAccMPI);
529 if (isChimeric) { MPIWroteAccnos = true; }
535 if((i+1) % 100 == 0){ cout << "Processing sequence: " << (i+1) << endl; m->mothurOutJustToLog("Processing sequence: " + toString(i+1) + "\n"); }
538 if(num % 100 != 0){ cout << "Processing sequence: " << num << endl; m->mothurOutJustToLog("Processing sequence: " + toString(num) + "\n"); }
543 catch(exception& e) {
544 m->errorOut(e, "ChimeraPintailCommand", "driverMPI");
550 /**************************************************************************************************/
552 int ChimeraPintailCommand::createProcesses(string outputFileName, string filename, string accnos) {
554 #if defined (__APPLE__) || (__MACH__) || (linux) || (__linux)
556 // processIDS.resize(0);
558 //loop through and create all the processes you want
559 while (process != processors) {
563 processIDS.push_back(pid); //create map from line number to pid so you can append files in correct order later
566 driver(lines[process], outputFileName + toString(getpid()) + ".temp", filename, accnos + toString(getpid()) + ".temp");
568 }else { m->mothurOut("unable to spawn the necessary processes."); m->mothurOutEndLine(); exit(0); }
571 //force parent to wait until all the processes are done
572 for (int i=0;i<processors;i++) {
573 int temp = processIDS[i];
580 catch(exception& e) {
581 m->errorOut(e, "ChimeraPintailCommand", "createProcesses");
586 /**************************************************************************************************/