X-Git-Url: https://git.donarmstrong.com/?a=blobdiff_plain;f=chimeraslayercommand.cpp;h=0958630862ebf3e61c9d2ec23e5618ddaa868ef1;hb=c47e480b743d1c242b8c527b6d12f992c68b8c2c;hp=cf219951f08e314f57ebb2e6c27985d3f5409167;hpb=77ac47e1ea0b5a0c6f55eb25e6bc7d7494ed3ad2;p=mothur.git diff --git a/chimeraslayercommand.cpp b/chimeraslayercommand.cpp index cf21995..0958630 100644 --- a/chimeraslayercommand.cpp +++ b/chimeraslayercommand.cpp @@ -8,8 +8,8 @@ */ #include "chimeraslayercommand.h" -#include "chimeraslayer.h" #include "deconvolutecommand.h" +#include "referencedb.h" //********************************************************************************************************************** vector ChimeraSlayerCommand::setParameters(){ @@ -23,21 +23,23 @@ vector ChimeraSlayerCommand::setParameters(){ CommandParameter pmismatch("mismatch", "Number", "", "-4.0", "", "", "",false,false); parameters.push_back(pmismatch); CommandParameter pminsim("minsim", "Number", "", "90", "", "", "",false,false); parameters.push_back(pminsim); CommandParameter pmincov("mincov", "Number", "", "70", "", "", "",false,false); parameters.push_back(pmincov); - CommandParameter pminsnp("minsnp", "Number", "", "100", "", "", "",false,false); parameters.push_back(pminsnp); + CommandParameter pminsnp("minsnp", "Number", "", "10", "", "", "",false,false); parameters.push_back(pminsnp); CommandParameter pminbs("minbs", "Number", "", "90", "", "", "",false,false); parameters.push_back(pminbs); - CommandParameter psearch("search", "Multiple", "kmer-blast-distance", "distance", "", "", "",false,false); parameters.push_back(psearch); + CommandParameter psearch("search", "Multiple", "kmer-blast", "blast", "", "", "",false,false); parameters.push_back(psearch); CommandParameter pprocessors("processors", "Number", "", "1", "", "", "",false,false); parameters.push_back(pprocessors); CommandParameter prealign("realign", "Boolean", "", "T", "", "", "",false,false); parameters.push_back(prealign); CommandParameter ptrim("trim", "Boolean", "", "F", "", "", "",false,false); parameters.push_back(ptrim); CommandParameter psplit("split", "Boolean", "", "F", "", "", "",false,false); parameters.push_back(psplit); CommandParameter pnumwanted("numwanted", "Number", "", "15", "", "", "",false,false); parameters.push_back(pnumwanted); - CommandParameter piters("iters", "Number", "", "100", "", "", "",false,false); parameters.push_back(piters); + CommandParameter piters("iters", "Number", "", "1000", "", "", "",false,false); parameters.push_back(piters); CommandParameter pdivergence("divergence", "Number", "", "1.007", "", "", "",false,false); parameters.push_back(pdivergence); CommandParameter pparents("parents", "Number", "", "3", "", "", "",false,false); parameters.push_back(pparents); CommandParameter pincrement("increment", "Number", "", "5", "", "", "",false,false); parameters.push_back(pincrement); + CommandParameter pblastlocation("blastlocation", "String", "", "", "", "", "",false,false); parameters.push_back(pblastlocation); CommandParameter pinputdir("inputdir", "String", "", "", "", "", "",false,false); parameters.push_back(pinputdir); CommandParameter poutputdir("outputdir", "String", "", "", "", "", "",false,false); parameters.push_back(poutputdir); - + CommandParameter psave("save", "Boolean", "", "F", "", "", "",false,false); parameters.push_back(psave); + vector myArray; for (int i = 0; i < parameters.size(); i++) { myArray.push_back(parameters[i].name); } return myArray; @@ -53,9 +55,9 @@ string ChimeraSlayerCommand::getHelpString(){ string helpString = ""; helpString += "The chimera.slayer command reads a fastafile and referencefile and outputs potentially chimeric sequences.\n"; helpString += "This command was modeled after the chimeraSlayer written by the Broad Institute.\n"; - helpString += "The chimera.slayer command parameters are fasta, name, template, processors, trim, ksize, window, match, mismatch, divergence. minsim, mincov, minbs, minsnp, parents, search, iters, increment, numwanted, and realign.\n"; + helpString += "The chimera.slayer command parameters are fasta, name, template, processors, trim, ksize, window, match, mismatch, divergence. minsim, mincov, minbs, minsnp, parents, search, iters, increment, numwanted, blastlocation and realign.\n"; helpString += "The fasta parameter allows you to enter the fasta file containing your potentially chimeric sequences, and is required, unless you have a valid current fasta file. \n"; - helpString += "The name parameter allows you to provide a name file, if you are using template=self. \n"; + helpString += "The name parameter allows you to provide a name file, if you are using reference=self. \n"; helpString += "You may enter multiple fasta files by separating their names with dashes. ie. fasta=abrecovery.fasta-amazon.fasta \n"; helpString += "The reference parameter allows you to enter a reference file containing known non-chimeric sequences, and is required. You may also set template=self, in this case the abundant sequences will be used as potential parents. \n"; helpString += "The processors parameter allows you to specify how many processors you would like to use. The default is 1. \n"; @@ -72,13 +74,15 @@ string ChimeraSlayerCommand::getHelpString(){ helpString += "The parents parameter allows you to select the number of potential parents to investigate from the numwanted best matches after rating them, default is 3. \n"; helpString += "The mismatch parameter allows you to penalize mismatched bases in blast search, default is -4. \n"; helpString += "The divergence parameter allows you to set a cutoff for chimera determination, default is 1.007. \n"; - helpString += "The iters parameter allows you to specify the number of bootstrap iters to do with the chimeraslayer method, default=100.\n"; + helpString += "The iters parameter allows you to specify the number of bootstrap iters to do with the chimeraslayer method, default=1000.\n"; helpString += "The minsim parameter allows you to specify a minimum similarity with the parent fragments, default=90. \n"; helpString += "The mincov parameter allows you to specify minimum coverage by closest matches found in template. Default is 70, meaning 70%. \n"; helpString += "The minbs parameter allows you to specify minimum bootstrap support for calling a sequence chimeric. Default is 90, meaning 90%. \n"; - helpString += "The minsnp parameter allows you to specify percent of SNPs to sample on each side of breakpoint for computing bootstrap support (default: 100) \n"; - helpString += "The search parameter allows you to specify search method for finding the closest parent. Choices are distance, blast, and kmer, default distance. \n"; + helpString += "The minsnp parameter allows you to specify percent of SNPs to sample on each side of breakpoint for computing bootstrap support (default: 10) \n"; + helpString += "The search parameter allows you to specify search method for finding the closest parent. Choices are blast, and kmer, default blast. \n"; helpString += "The realign parameter allows you to realign the query to the potential parents. Choices are true or false, default true. \n"; + helpString += "The blastlocation parameter allows you to specify the location of your blast executable. By default mothur will look in ./blast/bin relative to mothur's executable. \n"; + helpString += "If the save parameter is set to true the reference sequences will be saved in memory, to clear them later you can use the clear.memory command. Default=f."; helpString += "The chimera.slayer command should be in the following format: \n"; helpString += "chimera.slayer(fasta=yourFastaFile, reference=yourTemplate, search=yourSearch) \n"; helpString += "Example: chimera.slayer(fasta=AD.align, reference=core_set_aligned.imputed.fasta, search=kmer) \n"; @@ -109,9 +113,11 @@ ChimeraSlayerCommand::ChimeraSlayerCommand(){ ChimeraSlayerCommand::ChimeraSlayerCommand(string option) { try { abort = false; calledHelp = false; + ReferenceDB* rdb = ReferenceDB::getInstance(); //allow user to run help if(option == "help") { help(); abort = true; calledHelp = true; } + else if(option == "citation") { citation(); abort = true; calledHelp = true;} else { vector myArray = setParameters(); @@ -204,6 +210,8 @@ ChimeraSlayerCommand::ChimeraSlayerCommand(string option) { //erase from file list fastaFileNames.erase(fastaFileNames.begin()+i); i--; + }else { + m->setFastaFile(fastaFileNames[i]); } } } @@ -278,6 +286,8 @@ ChimeraSlayerCommand::ChimeraSlayerCommand(string option) { //erase from file list nameFileNames.erase(nameFileNames.begin()+i); i--; + }else { + m->setNameFile(nameFileNames[i]); } } } @@ -291,12 +301,29 @@ ChimeraSlayerCommand::ChimeraSlayerCommand(string option) { //if the user changes the output directory command factory will send this info to us in the output parameter outputDir = validParameter.validFile(parameters, "outputdir", false); if (outputDir == "not found"){ outputDir = ""; } + string temp = validParameter.validFile(parameters, "processors", false); if (temp == "not found"){ temp = m->getProcessors(); } + m->setProcessors(temp); + convert(temp, processors); + + temp = validParameter.validFile(parameters, "save", false); if (temp == "not found"){ temp = "f"; } + save = m->isTrue(temp); + rdb->save = save; + if (save) { //clear out old references + rdb->clearMemory(); + } string path; it = parameters.find("reference"); //user has given a template file if(it != parameters.end()){ - if (it->second == "self") { templatefile = "self"; } + if (it->second == "self") { + templatefile = "self"; + if (save) { + m->mothurOut("[WARNING]: You can't save reference=self, ignoring save."); + m->mothurOutEndLine(); + save = false; + } + } else { path = m->hasPath(it->second); //if the user has not given a path then, add inputdir. else leave path alone. @@ -304,13 +331,34 @@ ChimeraSlayerCommand::ChimeraSlayerCommand(string option) { templatefile = validParameter.validFile(parameters, "reference", true); if (templatefile == "not open") { abort = true; } - else if (templatefile == "not found") { templatefile = ""; m->mothurOut("reference is a required parameter for the chimera.slayer command."); m->mothurOutEndLine(); abort = true; } + else if (templatefile == "not found") { //check for saved reference sequences + if (rdb->referenceSeqs.size() != 0) { + templatefile = "saved"; + }else { + m->mothurOut("[ERROR]: You don't have any saved reference sequences and the reference parameter is a required."); + m->mothurOutEndLine(); + abort = true; + } + }else { if (save) { rdb->setSavedReference(templatefile); } } } + }else if (hasName) { templatefile = "self"; + if (save) { + m->mothurOut("[WARNING]: You can't save reference=self, ignoring save."); + m->mothurOutEndLine(); + save = false; + } + } + else { + if (rdb->referenceSeqs.size() != 0) { + templatefile = "saved"; + }else { + m->mothurOut("[ERROR]: You don't have any saved reference sequences and the reference parameter is a required."); + m->mothurOutEndLine(); + templatefile = ""; abort = true; + } } - string temp = validParameter.validFile(parameters, "processors", false); if (temp == "not found"){ temp = m->getProcessors(); } - m->setProcessors(temp); - convert(temp, processors); + temp = validParameter.validFile(parameters, "ksize", false); if (temp == "not found") { temp = "7"; } convert(temp, ksize); @@ -336,13 +384,13 @@ ChimeraSlayerCommand::ChimeraSlayerCommand(string option) { temp = validParameter.validFile(parameters, "minbs", false); if (temp == "not found") { temp = "90"; } convert(temp, minBS); - temp = validParameter.validFile(parameters, "minsnp", false); if (temp == "not found") { temp = "100"; } + temp = validParameter.validFile(parameters, "minsnp", false); if (temp == "not found") { temp = "10"; } convert(temp, minSNP); temp = validParameter.validFile(parameters, "parents", false); if (temp == "not found") { temp = "3"; } convert(temp, parents); - temp = validParameter.validFile(parameters, "realign", true); if (temp == "not found") { temp = "t"; } + temp = validParameter.validFile(parameters, "realign", false); if (temp == "not found") { temp = "t"; } realign = m->isTrue(temp); temp = validParameter.validFile(parameters, "trim", false); if (temp == "not found") { temp = "f"; } @@ -351,9 +399,9 @@ ChimeraSlayerCommand::ChimeraSlayerCommand(string option) { temp = validParameter.validFile(parameters, "split", false); if (temp == "not found") { temp = "f"; } trimera = m->isTrue(temp); - search = validParameter.validFile(parameters, "search", false); if (search == "not found") { search = "distance"; } + search = validParameter.validFile(parameters, "search", false); if (search == "not found") { search = "blast"; } - temp = validParameter.validFile(parameters, "iters", false); if (temp == "not found") { temp = "100"; } + temp = validParameter.validFile(parameters, "iters", false); if (temp == "not found") { temp = "1000"; } convert(temp, iters); temp = validParameter.validFile(parameters, "increment", false); if (temp == "not found") { temp = "5"; } @@ -361,8 +409,47 @@ ChimeraSlayerCommand::ChimeraSlayerCommand(string option) { temp = validParameter.validFile(parameters, "numwanted", false); if (temp == "not found") { temp = "15"; } convert(temp, numwanted); + + blastlocation = validParameter.validFile(parameters, "blastlocation", false); + if (blastlocation == "not found") { blastlocation = ""; } + else { + //add / to name if needed + string lastChar = blastlocation.substr(blastlocation.length()-1); +#if defined (__APPLE__) || (__MACH__) || (linux) || (__linux) + if (lastChar != "/") { blastlocation += "/"; } +#else + if (lastChar != "\\") { blastlocation += "\\"; } +#endif + blastlocation = m->getFullPathName(blastlocation); + string formatdbCommand = ""; +#if defined (__APPLE__) || (__MACH__) || (linux) || (__linux) + formatdbCommand = blastlocation + "formatdb"; +#else + formatdbCommand = blastlocation + "formatdb.exe"; +#endif + + //test to make sure formatdb exists + ifstream in; + formatdbCommand = m->getFullPathName(formatdbCommand); + int ableToOpen = m->openInputFile(formatdbCommand, in, "no error"); in.close(); + if(ableToOpen == 1) { m->mothurOut("[ERROR]: " + formatdbCommand + " file does not exist. mothur requires formatdb.exe to run chimera.slayer."); m->mothurOutEndLine(); abort = true; } + + string blastCommand = ""; +#if defined (__APPLE__) || (__MACH__) || (linux) || (__linux) + blastCommand = blastlocation + "megablast"; +#else + blastCommand = blastlocation + "megablast.exe"; +#endif + //test to make sure formatdb exists + ifstream in2; + blastCommand = m->getFullPathName(blastCommand); + ableToOpen = m->openInputFile(blastCommand, in2, "no error"); in2.close(); + if(ableToOpen == 1) { m->mothurOut("[ERROR]: " + blastCommand + " file does not exist. mothur requires blastall.exe to run chimera.slayer."); m->mothurOutEndLine(); abort = true; } + } - if ((search != "distance") && (search != "blast") && (search != "kmer")) { m->mothurOut(search + " is not a valid search."); m->mothurOutEndLine(); abort = true; } + if ((search != "blast") && (search != "kmer")) { m->mothurOut(search + " is not a valid search."); m->mothurOutEndLine(); abort = true; } + + if (hasName && (templatefile != "self")) { m->mothurOut("You have provided a namefile and the reference parameter is not set to self. I am not sure what reference you are trying to use, aborting."); m->mothurOutEndLine(); abort=true; } } } catch(exception& e) { @@ -374,18 +461,15 @@ ChimeraSlayerCommand::ChimeraSlayerCommand(string option) { int ChimeraSlayerCommand::execute(){ try{ - if (abort == true) { if (calledHelp) { return 0; } return 2; } - + for (int s = 0; s < fastaFileNames.size(); s++) { m->mothurOut("Checking sequences from " + fastaFileNames[s] + " ..." ); m->mothurOutEndLine(); int start = time(NULL); - if (templatefile != "self") { //you want to run slayer with a refernce template - chimera = new ChimeraSlayer(fastaFileNames[s], templatefile, trim, search, ksize, match, mismatch, window, divR, minSimilarity, minCoverage, minBS, minSNP, parents, iters, increment, numwanted, realign); - }else { + if (templatefile == "self") { if (processors != 1) { m->mothurOut("When using template=self, mothur can only use 1 processor, continuing."); m->mothurOutEndLine(); processors = 1; } string nameFile = ""; if (nameFileNames.size() != 0) { //you provided a namefile and we don't need to create one @@ -397,7 +481,7 @@ int ChimeraSlayerCommand::execute(){ string inputString = "fasta=" + fastaFileNames[s]; m->mothurOut("/******************************************/"); m->mothurOutEndLine(); m->mothurOut("Running command: unique.seqs(" + inputString + ")"); m->mothurOutEndLine(); - + Command* uniqueCommand = new DeconvoluteCommand(inputString); uniqueCommand->execute(); @@ -413,20 +497,26 @@ int ChimeraSlayerCommand::execute(){ //sort fastafile by abundance, returns new sorted fastafile name m->mothurOut("Sorting fastafile according to abundance..."); cout.flush(); - map priority = sortFastaFile(fastaFileNames[s], nameFile); + priority = sortFastaFile(fastaFileNames[s], nameFile); m->mothurOut("Done."); m->mothurOutEndLine(); - if (m->control_pressed) { for (int j = 0; j < outputNames.size(); j++) { remove(outputNames[j].c_str()); } return 0; } - - chimera = new ChimeraSlayer(fastaFileNames[s], templatefile, trim, priority, search, ksize, match, mismatch, window, divR, minSimilarity, minCoverage, minBS, minSNP, parents, iters, increment, numwanted, realign); + if (m->control_pressed) { for (int j = 0; j < outputNames.size(); j++) { m->mothurRemove(outputNames[j]); } return 0; } } - + if (outputDir == "") { outputDir = m->hasPath(fastaFileNames[s]); }//if user entered a file with a path then preserve it string outputFileName = outputDir + m->getRootName(m->getSimpleName(fastaFileNames[s])) + "slayer.chimera"; string accnosFileName = outputDir + m->getRootName(m->getSimpleName(fastaFileNames[s])) + "slayer.accnos"; string trimFastaFileName = outputDir + m->getRootName(m->getSimpleName(fastaFileNames[s])) + "slayer.fasta"; - if (m->control_pressed) { delete chimera; for (int j = 0; j < outputNames.size(); j++) { remove(outputNames[j].c_str()); } return 0; } + //create chimera here if you are mac or linux because fork will copy for you. Create in create processes instead if you are windows. + #if defined (__APPLE__) || (__MACH__) || (linux) || (__linux) || USE_MPI + if (templatefile != "self") { //you want to run slayer with a reference template + chimera = new ChimeraSlayer(fastaFileNames[s], templatefile, trim, search, ksize, match, mismatch, window, divR, minSimilarity, minCoverage, minBS, minSNP, parents, iters, increment, numwanted, realign, blastlocation, rand()); + }else { + chimera = new ChimeraSlayer(fastaFileNames[s], templatefile, trim, priority, search, ksize, match, mismatch, window, divR, minSimilarity, minCoverage, minBS, minSNP, parents, iters, increment, numwanted, realign, blastlocation, rand()); + } + + if (m->control_pressed) { delete chimera; for (int j = 0; j < outputNames.size(); j++) { m->mothurRemove(outputNames[j]); } return 0; } if (chimera->getUnaligned()) { m->mothurOut("Your template sequences are different lengths, please correct."); m->mothurOutEndLine(); @@ -434,11 +524,29 @@ int ChimeraSlayerCommand::execute(){ return 0; } templateSeqsLength = chimera->getLength(); + #else + if (processors == 1) { + if (templatefile != "self") { //you want to run slayer with a reference template + chimera = new ChimeraSlayer(fastaFileNames[s], templatefile, trim, search, ksize, match, mismatch, window, divR, minSimilarity, minCoverage, minBS, minSNP, parents, iters, increment, numwanted, realign, blastlocation, rand()); + }else { + chimera = new ChimeraSlayer(fastaFileNames[s], templatefile, trim, priority, search, ksize, match, mismatch, window, divR, minSimilarity, minCoverage, minBS, minSNP, parents, iters, increment, numwanted, realign, blastlocation, rand()); + } + + if (m->control_pressed) { delete chimera; for (int j = 0; j < outputNames.size(); j++) { m->mothurRemove(outputNames[j]); } return 0; } + + if (chimera->getUnaligned()) { + m->mothurOut("Your template sequences are different lengths, please correct."); m->mothurOutEndLine(); + delete chimera; + return 0; + } + templateSeqsLength = chimera->getLength(); + } + #endif #ifdef USE_MPI int pid, numSeqsPerProcessor; int tag = 2001; - vector MPIPos; + vector MPIPos; MPI_Status status; MPI_Comm_rank(MPI_COMM_WORLD, &pid); //find out who we are @@ -469,7 +577,7 @@ int ChimeraSlayerCommand::execute(){ MPI_File_open(MPI_COMM_WORLD, outAccnosFilename, outMode, MPI_INFO_NULL, &outMPIAccnos); if (trim) { MPI_File_open(MPI_COMM_WORLD, outFastaFilename, outMode, MPI_INFO_NULL, &outMPIFasta); } - if (m->control_pressed) { outputTypes.clear(); MPI_File_close(&inMPI); MPI_File_close(&outMPI); if (trim) { MPI_File_close(&outMPIFasta); } MPI_File_close(&outMPIAccnos); for (int j = 0; j < outputNames.size(); j++) { remove(outputNames[j].c_str()); } delete chimera; return 0; } + if (m->control_pressed) { outputTypes.clear(); MPI_File_close(&inMPI); MPI_File_close(&outMPI); if (trim) { MPI_File_close(&outMPIFasta); } MPI_File_close(&outMPIAccnos); for (int j = 0; j < outputNames.size(); j++) { m->mothurRemove(outputNames[j]); } delete chimera; return 0; } if (pid == 0) { //you are the root process m->mothurOutEndLine(); @@ -508,7 +616,17 @@ int ChimeraSlayerCommand::execute(){ //do your part driverMPI(startIndex, numSeqsPerProcessor, inMPI, outMPI, outMPIAccnos, outMPIFasta, MPIPos); - if (m->control_pressed) { outputTypes.clear(); MPI_File_close(&inMPI); MPI_File_close(&outMPI); if (trim) { MPI_File_close(&outMPIFasta); } MPI_File_close(&outMPIAccnos); for (int j = 0; j < outputNames.size(); j++) { remove(outputNames[j].c_str()); } remove(outputFileName.c_str()); remove(accnosFileName.c_str()); delete chimera; return 0; } + int numNoParents = chimera->getNumNoParents(); + int temp; + for(int i = 1; i < processors; i++) { + MPI_Recv(&temp, 1, MPI_INT, 1, tag, MPI_COMM_WORLD, &status); + numNoParents += temp; + } + + + if (numSeqs == numNoParents) { m->mothurOut("[WARNING]: megablast returned 0 potential parents for all your sequences. This could be due to formatdb.exe not being setup properly, please check formatdb.log for errors."); m->mothurOutEndLine(); } + + if (m->control_pressed) { outputTypes.clear(); MPI_File_close(&inMPI); MPI_File_close(&outMPI); if (trim) { MPI_File_close(&outMPIFasta); } MPI_File_close(&outMPIAccnos); for (int j = 0; j < outputNames.size(); j++) { m->mothurRemove(outputNames[j]); } m->mothurRemove(outputFileName); m->mothurRemove(accnosFileName); delete chimera; return 0; } }else{ //you are a child process if (templatefile != "self") { //if template=self we can only use 1 processor @@ -523,8 +641,11 @@ int ChimeraSlayerCommand::execute(){ //do your part driverMPI(startIndex, numSeqsPerProcessor, inMPI, outMPI, outMPIAccnos, outMPIFasta, MPIPos); + + int numNoParents = chimera->getNumNoParents(); + MPI_Send(&numNoParents, 1, MPI_INT, 0, tag, MPI_COMM_WORLD); - if (m->control_pressed) { outputTypes.clear(); MPI_File_close(&inMPI); MPI_File_close(&outMPI); if (trim) { MPI_File_close(&outMPIFasta); } MPI_File_close(&outMPIAccnos); for (int j = 0; j < outputNames.size(); j++) { remove(outputNames[j].c_str()); } delete chimera; return 0; } + if (m->control_pressed) { outputTypes.clear(); MPI_File_close(&inMPI); MPI_File_close(&outMPI); if (trim) { MPI_File_close(&outMPIFasta); } MPI_File_close(&outMPIAccnos); for (int j = 0; j < outputNames.size(); j++) { m->mothurRemove(outputNames[j]); } delete chimera; return 0; } } } @@ -537,72 +658,67 @@ int ChimeraSlayerCommand::execute(){ MPI_Barrier(MPI_COMM_WORLD); //make everyone wait - just in case #else - ofstream outHeader; - string tempHeader = outputDir + m->getRootName(m->getSimpleName(fastaFileNames[s])) + "slayer.chimeras.tempHeader"; - m->openOutputFile(tempHeader, outHeader); - - chimera->printHeader(outHeader); - outHeader.close(); + //break up file + vector positions; +#if defined (__APPLE__) || (__MACH__) || (linux) || (__linux) + positions = m->divideFile(fastaFileNames[s], processors); + for (int i = 0; i < (positions.size()-1); i++) { lines.push_back(new linePair(positions[i], positions[(i+1)])); } +#else + if (processors == 1) { + lines.push_back(new linePair(0, 1000)); + }else { + positions = m->setFilePosFasta(fastaFileNames[s], numSeqs); + + //figure out how many sequences you have to process + int numSeqsPerProcessor = numSeqs / processors; + for (int i = 0; i < processors; i++) { + int startIndex = i * numSeqsPerProcessor; + if(i == (processors - 1)){ numSeqsPerProcessor = numSeqs - i * numSeqsPerProcessor; } + lines.push_back(new linePair(positions[startIndex], numSeqsPerProcessor)); + } + } +#endif - vector positions = m->divideFile(fastaFileNames[s], processors); + if(processors == 1){ + numSeqs = driver(lines[0], outputFileName, fastaFileNames[s], accnosFileName, trimFastaFileName); - for (int i = 0; i < (positions.size()-1); i++) { - lines.push_back(new linePair(positions[i], positions[(i+1)])); - } - - //break up file - #if defined (__APPLE__) || (__MACH__) || (linux) || (__linux) - if(processors == 1){ - numSeqs = driver(lines[0], outputFileName, fastaFileNames[s], accnosFileName, trimFastaFileName); - - if (m->control_pressed) { outputTypes.clear(); if (trim) { remove(trimFastaFileName.c_str()); } 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]; } lines.clear(); delete chimera; return 0; } - - }else{ - processIDS.resize(0); - - numSeqs = createProcesses(outputFileName, fastaFileNames[s], accnosFileName, trimFastaFileName); + int numNoParents = chimera->getNumNoParents(); + if (numNoParents == numSeqs) { m->mothurOut("[WARNING]: megablast returned 0 potential parents for all your sequences. This could be due to formatdb.exe not being setup properly, please check formatdb.log for errors."); m->mothurOutEndLine(); } - rename((outputFileName + toString(processIDS[0]) + ".temp").c_str(), outputFileName.c_str()); - rename((accnosFileName + toString(processIDS[0]) + ".temp").c_str(), accnosFileName.c_str()); - if (trim) { rename((trimFastaFileName + toString(processIDS[0]) + ".temp").c_str(), trimFastaFileName.c_str()); } - - //append output files - for(int i=1;iappendFiles((outputFileName + toString(processIDS[i]) + ".temp"), outputFileName); - remove((outputFileName + toString(processIDS[i]) + ".temp").c_str()); - } + if (m->control_pressed) { outputTypes.clear(); if (trim) { m->mothurRemove(trimFastaFileName); } m->mothurRemove(outputFileName); m->mothurRemove(accnosFileName); for (int j = 0; j < outputNames.size(); j++) { m->mothurRemove(outputNames[j]); } for (int i = 0; i < lines.size(); i++) { delete lines[i]; } lines.clear(); delete chimera; return 0; } + + }else{ + processIDS.resize(0); + + numSeqs = createProcesses(outputFileName, fastaFileNames[s], accnosFileName, trimFastaFileName); + + rename((outputFileName + toString(processIDS[0]) + ".temp").c_str(), outputFileName.c_str()); + rename((accnosFileName + toString(processIDS[0]) + ".temp").c_str(), accnosFileName.c_str()); + if (trim) { rename((trimFastaFileName + toString(processIDS[0]) + ".temp").c_str(), trimFastaFileName.c_str()); } - //append output files - for(int i=1;iappendFiles((accnosFileName + toString(processIDS[i]) + ".temp"), accnosFileName); - remove((accnosFileName + toString(processIDS[i]) + ".temp").c_str()); - } + //append output files + for(int i=1;iappendFiles((outputFileName + toString(processIDS[i]) + ".temp"), outputFileName); + m->mothurRemove((outputFileName + toString(processIDS[i]) + ".temp")); + + m->appendFiles((accnosFileName + toString(processIDS[i]) + ".temp"), accnosFileName); + m->mothurRemove((accnosFileName + toString(processIDS[i]) + ".temp")); if (trim) { - for(int i=1;iappendFiles((trimFastaFileName + toString(processIDS[i]) + ".temp"), trimFastaFileName); - remove((trimFastaFileName + toString(processIDS[i]) + ".temp").c_str()); - } + m->appendFiles((trimFastaFileName + toString(processIDS[i]) + ".temp"), trimFastaFileName); + m->mothurRemove((trimFastaFileName + toString(processIDS[i]) + ".temp")); } - - if (m->control_pressed) { outputTypes.clear(); if (trim) { remove(trimFastaFileName.c_str()); } remove(outputFileName.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]; } lines.clear(); delete chimera; return 0; } } - - #else - numSeqs = driver(lines[0], outputFileName, fastaFileNames[s], accnosFileName, trimFastaFileName); - - if (m->control_pressed) { outputTypes.clear(); if (trim) { remove(trimFastaFileName.c_str()); } 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]; } lines.clear(); delete chimera; return 0; } - #endif - - m->appendFiles(outputFileName, tempHeader); - - remove(outputFileName.c_str()); - rename(tempHeader.c_str(), outputFileName.c_str()); + if (m->control_pressed) { outputTypes.clear(); if (trim) { m->mothurRemove(trimFastaFileName); } m->mothurRemove(outputFileName); m->mothurRemove(accnosFileName); for (int j = 0; j < outputNames.size(); j++) { m->mothurRemove(outputNames[j]); } for (int i = 0; i < lines.size(); i++) { delete lines[i]; } lines.clear(); delete chimera; return 0; } + } + #endif - delete chimera; + #if defined (__APPLE__) || (__MACH__) || (linux) || (__linux) || USE_MPI + delete chimera; + #endif for (int i = 0; i < lines.size(); i++) { delete lines[i]; } lines.clear(); @@ -657,6 +773,8 @@ int ChimeraSlayerCommand::driver(linePair* filePos, string outputFName, string f m->openInputFile(filename, inFASTA); inFASTA.seekg(filePos->start); + + if (filePos->start == 0) { chimera->printHeader(out); } bool done = false; int count = 0; @@ -669,7 +787,6 @@ int ChimeraSlayerCommand::driver(linePair* filePos, string outputFName, string f string candidateAligned = candidateSeq->getAligned(); if (candidateSeq->getName() != "") { //incase there is a commented sequence at the end of a file - if (candidateSeq->getAligned().length() != templateSeqsLength) { m->mothurOut(candidateSeq->getName() + " is not the same length as the template sequences. Skipping."); m->mothurOutEndLine(); }else{ @@ -715,15 +832,15 @@ int ChimeraSlayerCommand::driver(linePair* filePos, string outputFName, string f data_results rightResults = chimera->getResults(); //if either piece is chimeric then report - Sequence* trimmed = chimera->print(out, out2, leftResults, rightResults); - if (trim) { trimmed->printSequence(out3); delete trimmed; } + Sequence trimmed = chimera->print(out, out2, leftResults, rightResults); + if (trim) { trimmed.printSequence(out3); } delete left; delete right; }else { //already chimeric //print results - Sequence* trimmed = chimera->print(out, out2); - if (trim) { trimmed->printSequence(out3); delete trimmed; } + Sequence trimmed = chimera->print(out, out2); + if (trim) { trimmed.printSequence(out3); } } @@ -732,8 +849,7 @@ int ChimeraSlayerCommand::driver(linePair* filePos, string outputFName, string f } #if defined (__APPLE__) || (__MACH__) || (linux) || (__linux) - unsigned long int pos = inFASTA.tellg(); - //cout << candidateSeq->getName() << '\t' << pos << endl; + unsigned long long pos = inFASTA.tellg(); if ((pos == -1) || (pos >= filePos->end)) { break; } #else if (inFASTA.eof()) { break; } @@ -760,7 +876,7 @@ int ChimeraSlayerCommand::driver(linePair* filePos, string outputFName, string f } //********************************************************************************************************************** #ifdef USE_MPI -int ChimeraSlayerCommand::driverMPI(int start, int num, MPI_File& inMPI, MPI_File& outMPI, MPI_File& outAccMPI, MPI_File& outFastaMPI, vector& MPIPos){ +int ChimeraSlayerCommand::driverMPI(int start, int num, MPI_File& inMPI, MPI_File& outMPI, MPI_File& outAccMPI, MPI_File& outFastaMPI, vector& MPIPos){ try { MPI_Status status; int pid; @@ -832,10 +948,9 @@ int ChimeraSlayerCommand::driverMPI(int start, int num, MPI_File& inMPI, MPI_Fil data_results rightResults = chimera->getResults(); //if either piece is chimeric then report - Sequence* trimmed = chimera->print(outMPI, outAccMPI, leftResults, rightResults); + Sequence trimmed = chimera->print(outMPI, outAccMPI, leftResults, rightResults); if (trim) { - string outputString = ">" + trimmed->getName() + "\n" + trimmed->getAligned() + "\n"; - delete trimmed; + string outputString = ">" + trimmed.getName() + "\n" + trimmed.getAligned() + "\n"; //write to accnos file int length = outputString.length(); @@ -850,11 +965,10 @@ int ChimeraSlayerCommand::driverMPI(int start, int num, MPI_File& inMPI, MPI_Fil }else { //print results - Sequence* trimmed = chimera->print(outMPI, outAccMPI); + Sequence trimmed = chimera->print(outMPI, outAccMPI); if (trim) { - string outputString = ">" + trimmed->getName() + "\n" + trimmed->getAligned() + "\n"; - delete trimmed; + string outputString = ">" + trimmed.getName() + "\n" + trimmed.getAligned() + "\n"; //write to accnos file int length = outputString.length(); @@ -890,10 +1004,11 @@ int ChimeraSlayerCommand::driverMPI(int start, int num, MPI_File& inMPI, MPI_Fil int ChimeraSlayerCommand::createProcesses(string outputFileName, string filename, string accnos, string fasta) { try { -#if defined (__APPLE__) || (__MACH__) || (linux) || (__linux) int process = 0; int num = 0; + int numNoParents = 0; +#if defined (__APPLE__) || (__MACH__) || (linux) || (__linux) //loop through and create all the processes you want while (process != processors) { int pid = fork(); @@ -908,9 +1023,8 @@ int ChimeraSlayerCommand::createProcesses(string outputFileName, string filename ofstream out; string tempFile = outputFileName + toString(getpid()) + ".num.temp"; m->openOutputFile(tempFile, out); - out << num << endl; + out << num << '\t' << chimera->getNumNoParents() << endl; out.close(); - exit(0); }else { m->mothurOut("[ERROR]: unable to spawn the necessary processes."); m->mothurOutEndLine(); @@ -929,12 +1043,45 @@ int ChimeraSlayerCommand::createProcesses(string outputFileName, string filename ifstream in; string tempFile = outputFileName + toString(processIDS[i]) + ".num.temp"; m->openInputFile(tempFile, in); - if (!in.eof()) { int tempNum = 0; in >> tempNum; num += tempNum; } - in.close(); remove(tempFile.c_str()); + if (!in.eof()) { int tempNum = 0; int tempNumParents = 0; in >> tempNum >> tempNumParents; num += tempNum; numNoParents += tempNumParents; } + in.close(); m->mothurRemove(tempFile); } +#else + + ////////////////////////////////////////////////////////////////////////////////////////////////////// + //Windows version shared memory, so be careful when passing variables through the slayerData struct. + //Above fork() will clone, so memory is separate, but that's not the case with windows, + ////////////////////////////////////////////////////////////////////////////////////////////////////// + vector pDataArray; + DWORD dwThreadIdArray[processors]; + HANDLE hThreadArray[processors]; + + //Create processor worker threads. + for( int i=0; istart, lines[i]->end, ksize, match, mismatch, window, minSimilarity, minCoverage, minBS, minSNP, parents, iters, increment, numwanted, divR, priority, i); + pDataArray.push_back(tempslayer); + processIDS.push_back(i); + + //MySeqSumThreadFunction is in header. It must be global or static to work with the threads. + //default security attributes, thread function name, argument to thread function, use default creation flags, returns the thread identifier + hThreadArray[i] = CreateThread(NULL, 0, MySlayerThreadFunction, pDataArray[i], 0, &dwThreadIdArray[i]); + } + + //Wait until all threads have terminated. + WaitForMultipleObjects(processors, hThreadArray, TRUE, INFINITE); + + //Close all thread handles and free memory allocations. + for(int i=0; i < pDataArray.size(); i++){ + num += pDataArray[i]->count; + numNoParents += pDataArray[i]->numNoParents; + CloseHandle(hThreadArray[i]); + delete pDataArray[i]; + } +#endif + if (num == numNoParents) { m->mothurOut("[WARNING]: megablast returned 0 potential parents for all your sequences. This could be due to formatdb.exe not being setup properly, please check formatdb.log for errors."); m->mothurOutEndLine(); } return num; -#endif } catch(exception& e) { m->errorOut(e, "ChimeraSlayerCommand", "createProcesses");