X-Git-Url: https://git.donarmstrong.com/?a=blobdiff_plain;f=aligncommand.cpp;h=958f29b39213c3fda29e810af8832fef4afa83e1;hb=30b3ffcd1cfd08e7144ae721bb53e27eb3f7a5d1;hp=4b9dedc97c0b6003f3d4cb49f76de2bac5f2beab;hpb=0470f6d037aacb3563c3f7010708120a4a67d4e6;p=mothur.git diff --git a/aligncommand.cpp b/aligncommand.cpp index 4b9dedc..958f29b 100644 --- a/aligncommand.cpp +++ b/aligncommand.cpp @@ -12,23 +12,16 @@ * allows for a global alignment and not the local alignment provided by bLAst. Furthermore, it has the potential to * provide a better alignment because of the banding method employed by blast (I'm not sure about this). * - * to compile type: - * make - * - * for basic instructions on how to run nastPlus type: - * ./nastPlus */ #include "aligncommand.h" #include "sequence.hpp" -#include "alignment.hpp" #include "gotohoverlap.hpp" #include "needlemanoverlap.hpp" #include "blastalign.hpp" #include "noalign.hpp" -#include "database.hpp" #include "kmerdb.hpp" #include "suffixdb.hpp" #include "blastdb.hpp" @@ -38,112 +31,115 @@ //********************************************************************************************************************** + AlignCommand::AlignCommand(string option){ try { - globaldata = GlobalData::getInstance(); + // globaldata = GlobalData::getInstance(); abort = false; //allow user to run help if(option == "help") { help(); abort = true; } else { + //valid paramters for this command - string AlignArray[] = {"fasta","candidate","search","ksize","align","match","mismatch","gapopen","gapextend"}; + string AlignArray[] = {"template","candidate","search","ksize","align","match","mismatch","gapopen","gapextend", "processors"}; vector myArray (AlignArray, AlignArray+(sizeof(AlignArray)/sizeof(string))); - parser = new OptionParser(); - - parser->parse(option, parameters); delete parser; + OptionParser parser(option); + map parameters = parser.getParameters(); + + ValidParameters validParameter; - ValidParameters* validParameter = new ValidParameters(); - //check to make sure all parameters are valid for command - for (it = parameters.begin(); it != parameters.end(); it++) { - if (validParameter->isValidParameter(it->first, myArray, it->second) != true) { abort = true; } + for (map::iterator it = parameters.begin(); it != parameters.end(); it++) { + if (validParameter.isValidParameter(it->first, myArray, it->second) != true) { abort = true; } } //check for required parameters - templateFileName = validParameter->validFile(parameters, "fasta", true); - if (templateFileName == "not found") { cout << "fasta is a required parameter for the align.seqs command." << endl; abort = true; } + templateFileName = validParameter.validFile(parameters, "template", true); + if (templateFileName == "not found") { + mothurOut("template is a required parameter for the align.seqs command."); + mothurOutEndLine(); + abort = true; + } else if (templateFileName == "not open") { abort = true; } - else { globaldata->setFastaFile(templateFileName); } - - candidateFileName = validParameter->validFile(parameters, "candidate", true); - if (candidateFileName == "not found") { cout << "candidate is a required parameter for the align.seqs command." << endl; abort = true; } - else if (candidateFileName == "not open") { abort = true; } - else { - globaldata->setCandidateFile(candidateFileName); - openInputFile(candidateFileName, in); + + candidateFileName = validParameter.validFile(parameters, "candidate", true); + if (candidateFileName == "not found") { + mothurOut("candidate is a required parameter for the align.seqs command."); + mothurOutEndLine(); + abort = true; } + else if (candidateFileName == "not open") { abort = true; } - //check for optional parameter and set defaults // ...at some point should added some additional type checking... string temp; - temp = validParameter->validFile(parameters, "ksize", false); if (temp == "not found") { temp = "8"; } + temp = validParameter.validFile(parameters, "ksize", false); if (temp == "not found"){ temp = "8"; } convert(temp, kmerSize); - - temp = validParameter->validFile(parameters, "match", false); if (temp == "not found") { temp = "1.0"; } + + temp = validParameter.validFile(parameters, "match", false); if (temp == "not found"){ temp = "1.0"; } convert(temp, match); - - temp = validParameter->validFile(parameters, "mismatch", false); if (temp == "not found") { temp = "-1.0"; } + + temp = validParameter.validFile(parameters, "mismatch", false); if (temp == "not found"){ temp = "-1.0"; } convert(temp, misMatch); - - temp = validParameter->validFile(parameters, "gapopen", false); if (temp == "not found") { temp = "-1.0"; } + + temp = validParameter.validFile(parameters, "gapopen", false); if (temp == "not found"){ temp = "-2.0"; } convert(temp, gapOpen); - - temp = validParameter->validFile(parameters, "gapextend", false); if (temp == "not found") { temp = "-2.0"; } + + temp = validParameter.validFile(parameters, "gapextend", false); if (temp == "not found"){ temp = "-1.0"; } convert(temp, gapExtend); - - search = validParameter->validFile(parameters, "search", false); if (search == "not found") { search = "kmer"; } - align = validParameter->validFile(parameters, "align", false); if (align == "not found") { align = "needleman"; } - delete validParameter; + temp = validParameter.validFile(parameters, "processors", false); if (temp == "not found"){ temp = "1"; } + convert(temp, processors); + + search = validParameter.validFile(parameters, "search", false); if (search == "not found"){ search = "kmer"; } + + align = validParameter.validFile(parameters, "align", false); if (align == "not found"){ align = "needleman"; } } - + } catch(exception& e) { - cout << "Standard Error: " << e.what() << " has occurred in the AlignCommand class Function AlignCommand. Please contact Pat Schloss at pschloss@microbio.umass.edu." << "\n"; + errorOut(e, "AlignCommand", "AlignCommand"); exit(1); } - catch(...) { - cout << "An unknown error has occurred in the AlignCommand class function AlignCommand. Please contact Pat Schloss at pschloss@microbio.umass.edu." << "\n"; - exit(1); - } } //********************************************************************************************************************** -AlignCommand::~AlignCommand(){ - +AlignCommand::~AlignCommand(){ + + if (abort == false) { + for (int i = 0; i < lines.size(); i++) { delete lines[i]; } lines.clear(); + delete templateDB; + delete alignment; + } } + //********************************************************************************************************************** void AlignCommand::help(){ try { - cout << "The align.seqs command reads a file containing sequences and creates an alignment file and a report file." << "\n"; - cout << "The align.seqs command parameters are fasta, candidate, search, ksize, align, match, mismatch, gapopen and gapextend. " << "\n"; - cout << "The fasta and candidate parameters are required." << "\n"; - cout << "The search parameter allows you to specify the method to find most similar template. Your options are: suffix, kmer and blast. The default is kmer." << "\n"; - cout << "The align parameter allows you to specify the alignment method to use. Your options are: gotoh, needleman, blast and noalign. The default is needleman." << "\n"; - cout << "The ksize parameter allows you to specify the kmer size for finding most similar template to candidate. The default is 7." << "\n"; - cout << "The match parameter allows you to specify the bonus for having the same base. The default is 1.0." << "\n"; - cout << "The mistmatch parameter allows you to specify the penalty for having different bases. The default is -1.0." << "\n"; - cout << "The gapopen parameter allows you to specify the penalty for opening a gap in an alignment. The default is -1.0." << "\n"; - cout << "The gapextend parameter allows you to specify the penalty for extending a gap in an alignment. The default is -2.0." << "\n"; - cout << "The align.seqs command should be in the following format: " << "\n"; - cout << "align.seqs(fasta=yourTemplateFile, candidate=yourCandidateFile, align=yourAlignmentMethod, search=yourSearchmethod, ksize=yourKmerSize, match=yourMatchBonus, mismatch=yourMismatchpenalty, gapopen=yourGapopenPenalty, gapextend=yourGapExtendPenalty) " << "\n"; - cout << "Example align.seqs(candidate=candidate.fasta, fasta=core.filtered, align=kmer, search=gotoh, ksize=8, match=2.0, mismatch=3.0, gapopen=-2.0, gapextend=-1.0)" << "\n"; - cout << "Note: No spaces between parameter labels (i.e. fasta), '=' and parameters (i.e.yourFastaFile)." << "\n" << "\n"; + mothurOut("The align.seqs command reads a file containing sequences and creates an alignment file and a report file.\n"); + mothurOut("The align.seqs command parameters are template, candidate, search, ksize, align, match, mismatch, gapopen and gapextend.\n"); + mothurOut("The template and candidate parameters are required.\n"); + mothurOut("The search parameter allows you to specify the method to find most similar template. Your options are: suffix, kmer and blast. The default is kmer.\n"); + mothurOut("The align parameter allows you to specify the alignment method to use. Your options are: gotoh, needleman, blast and noalign. The default is needleman.\n"); + mothurOut("The ksize parameter allows you to specify the kmer size for finding most similar template to candidate. The default is 7.\n"); + mothurOut("The match parameter allows you to specify the bonus for having the same base. The default is 1.0.\n"); + mothurOut("The mistmatch parameter allows you to specify the penalty for having different bases. The default is -1.0.\n"); + mothurOut("The gapopen parameter allows you to specify the penalty for opening a gap in an alignment. The default is -1.0.\n"); + mothurOut("The gapextend parameter allows you to specify the penalty for extending a gap in an alignment. The default is -2.0.\n"); + mothurOut("The align.seqs command should be in the following format: \n"); + mothurOut("align.seqs(template=yourTemplateFile, candidate=yourCandidateFile, align=yourAlignmentMethod, search=yourSearchmethod, ksize=yourKmerSize, match=yourMatchBonus, mismatch=yourMismatchpenalty, gapopen=yourGapopenPenalty, gapextend=yourGapExtendPenalty) \n"); + mothurOut("Example align.seqs(candidate=candidate.fasta, template=core.filtered, align=kmer, search=gotoh, ksize=8, match=2.0, mismatch=3.0, gapopen=-2.0, gapextend=-1.0)\n"); + mothurOut("Note: No spaces between parameter labels (i.e. candidate), '=' and parameters (i.e.yourFastaFile).\n\n"); } catch(exception& e) { - cout << "Standard Error: " << e.what() << " has occurred in the AlignCommand class Function help. Please contact Pat Schloss at pschloss@microbio.umass.edu." << "\n"; + errorOut(e, "AlignCommand", "help"); exit(1); } - catch(...) { - cout << "An unknown error has occurred in the AlignCommand class function help. Please contact Pat Schloss at pschloss@microbio.umass.edu." << "\n"; - exit(1); - } } @@ -152,85 +148,244 @@ void AlignCommand::help(){ int AlignCommand::execute(){ try { if (abort == true) { return 0; } - - srand( (unsigned)time( NULL ) ); //needed to assign names to temporary files - Database* templateDB; if(search == "kmer") { templateDB = new KmerDB(templateFileName, kmerSize); } else if(search == "suffix") { templateDB = new SuffixDB(templateFileName); } else if(search == "blast") { templateDB = new BlastDB(templateFileName, gapOpen, gapExtend, match, misMatch); } else { - cout << search << " is not a valid search option. I will run the command using kmer, ksize=8." << endl; kmerSize = 8; + mothurOut(search + " is not a valid search option. I will run the command using kmer, ksize=8."); + mothurOutEndLine(); + kmerSize = 8; + templateDB = new KmerDB(templateFileName, kmerSize); } - - Alignment* alignment; - if(align == "gotoh") { alignment = new GotohOverlap(gapOpen, gapExtend, match, misMatch, 3000); } - else if(align == "needleman") { alignment = new NeedlemanOverlap(gapOpen, match, misMatch, 3000); } + + int longestBase = templateDB->getLongestBase(); + + if(align == "gotoh") { alignment = new GotohOverlap(gapOpen, gapExtend, match, misMatch, longestBase); } + else if(align == "needleman") { alignment = new NeedlemanOverlap(gapOpen, match, misMatch, longestBase); } else if(align == "blast") { alignment = new BlastAlignment(gapOpen, gapExtend, match, misMatch); } else if(align == "noalign") { alignment = new NoAlign(); } else { - cout << align << " is not a valid alignment option. I will run the command using needleman." << endl; - alignment = new NeedlemanOverlap(gapOpen, match, misMatch, 3000); + mothurOut(align + " is not a valid alignment option. I will run the command using needleman."); + mothurOutEndLine(); + alignment = new NeedlemanOverlap(gapOpen, match, misMatch, longestBase); + } + mothurOut("Aligning sequences..."); + mothurOutEndLine(); + + string alignFileName = candidateFileName.substr(0,candidateFileName.find_last_of(".")+1) + "align"; + string reportFileName = candidateFileName.substr(0,candidateFileName.find_last_of(".")+1) + "align.report"; + + int numFastaSeqs = 0; + int start = time(NULL); + +#if defined (__APPLE__) || (__MACH__) || (linux) || (__linux) + if(processors == 1){ + ifstream inFASTA; + openInputFile(candidateFileName, inFASTA); + numFastaSeqs=count(istreambuf_iterator(inFASTA),istreambuf_iterator(), '>'); + inFASTA.close(); + + lines.push_back(new linePair(0, numFastaSeqs)); + + driver(lines[0], alignFileName, reportFileName); + } + else{ + vector positions; + processIDS.resize(0); + + ifstream inFASTA; + openInputFile(candidateFileName, inFASTA); + + string input; + while(!inFASTA.eof()){ + input = getline(inFASTA); + if (input.length() != 0) { + if(input[0] == '>'){ int pos = inFASTA.tellg(); positions.push_back(pos - input.length() - 1); } + } + } + inFASTA.close(); + + numFastaSeqs = positions.size(); + + int numSeqsPerProcessor = numFastaSeqs / processors; + + for (int i = 0; i < processors; i++) { + int startPos = positions[ i * numSeqsPerProcessor ]; + if(i == processors - 1){ + numSeqsPerProcessor = numFastaSeqs - i * numSeqsPerProcessor; + } + lines.push_back(new linePair(startPos, numSeqsPerProcessor)); + } + createProcesses(alignFileName, reportFileName); + + rename((alignFileName + toString(processIDS[0]) + ".temp").c_str(), alignFileName.c_str()); + rename((reportFileName + toString(processIDS[0]) + ".temp").c_str(), reportFileName.c_str()); + + for(int i=1;i(in),istreambuf_iterator(), '>'); - in.seekg(0); - - string candidateAligngmentFName = candidateFileName.substr(0,candidateFileName.find_last_of(".")+1) + "align"; - ofstream candidateAlignmentFile; - openOutputFile(candidateAligngmentFName, candidateAlignmentFile); + appendReportFiles((reportFileName + toString(processIDS[i]) + ".temp"), reportFileName); + remove((reportFileName + toString(processIDS[i]) + ".temp").c_str()); + } + + } +#else + ifstream inFASTA; + openInputFile(candidateFileName, inFASTA); + numFastaSeqs=count(istreambuf_iterator(inFASTA),istreambuf_iterator(), '>'); + inFASTA.close(); + + lines.push_back(new linePair(0, numFastaSeqs)); + + driver(lines[0], alignFileName, reportFileName); +#endif + + mothurOut("It took " + toString(time(NULL) - start) + " secs to align " + toString(numFastaSeqs) + " sequences."); + mothurOutEndLine(); + mothurOutEndLine(); + + return 0; + } + catch(exception& e) { + errorOut(e, "AlignCommand", "execute"); + exit(1); + } +} - string candidateReportFName = candidateFileName.substr(0,candidateFileName.find_last_of(".")+1) + "align.report"; - NastReport report(candidateReportFName); +//********************************************************************************************************************** - cout << "We are going to align the " << numFastaSeqs << " sequences in " << candidateFileName << "..." << endl; - cout.flush(); - - int start = time(NULL); +int AlignCommand::driver(linePair* line, string alignFName, string reportFName){ + try { + ofstream alignmentFile; + openOutputFile(alignFName, alignmentFile); + NastReport report(reportFName); + + ifstream inFASTA; + openInputFile(candidateFileName, inFASTA); - for(int i=0;istart); - Sequence* candidateSeq = new Sequence(in); + for(int i=0;inumSeqs;i++){ + + Sequence* candidateSeq = new Sequence(inFASTA); report.setCandidate(candidateSeq); - - Sequence* templateSeq = templateDB->findClosestSequence(candidateSeq); + + Sequence temp = templateDB->findClosestSequence(candidateSeq); + Sequence* templateSeq = &temp; + report.setTemplate(templateSeq); report.setSearchParameters(search, templateDB->getSearchScore()); - Nast nast(alignment, candidateSeq, templateSeq); + report.setAlignmentParameters(align, alignment); - report.setNastParameters(nast); - candidateAlignmentFile << '>' << candidateSeq->getName() << '\n' << candidateSeq->getAligned() << endl; - candidateAlignmentFile.flush(); + report.setNastParameters(nast); - if((i+1) % 100 == 0){ - cout << "It has taken " << time(NULL) - start << " secs to align " << i+1 << " sequences" << endl; - } + alignmentFile << '>' << candidateSeq->getName() << '\n' << candidateSeq->getAligned() << endl; + report.print(); - + delete candidateSeq; } - cout << "It took " << time(NULL) - start << " secs to align " << numFastaSeqs << " sequences" << endl; - cout << endl; + alignmentFile.close(); + inFASTA.close(); + + return 1; + } + catch(exception& e) { + errorOut(e, "AlignCommand", "driver"); + exit(1); + } +} - delete templateDB; - delete alignment; +/**************************************************************************************************/ - - return 0; +void AlignCommand::createProcesses(string alignFileName, string reportFileName) { + try { +#if defined (__APPLE__) || (__MACH__) || (linux) || (__linux) + int process = 0; + // processIDS.resize(0); + + //loop through and create all the processes you want + while (process != processors) { + int pid = fork(); + + if (pid > 0) { + processIDS.push_back(pid); //create map from line number to pid so you can append files in correct order later + process++; + }else if (pid == 0){ + driver(lines[process], alignFileName + toString(getpid()) + ".temp", reportFileName + toString(getpid()) + ".temp"); + exit(0); + }else { mothurOut("unable to spawn the necessary processes."); mothurOutEndLine(); exit(0); } + } + + //force parent to wait until all the processes are done + for (int i=0;i