X-Git-Url: https://git.donarmstrong.com/?p=mothur.git;a=blobdiff_plain;f=pairwiseseqscommand.cpp;h=c9e5ecfdfda727427793e98aaed0270cdcd6e046;hp=733c47063bad878d8a54b061cc06e28fa2c7618c;hb=1a20e24ee786195ab0e1cccd4f5aede7a88f3f4e;hpb=df41d90a9611eab75d0516e2654ed8580f8df24c diff --git a/pairwiseseqscommand.cpp b/pairwiseseqscommand.cpp index 733c470..c9e5ecf 100644 --- a/pairwiseseqscommand.cpp +++ b/pairwiseseqscommand.cpp @@ -8,60 +8,83 @@ */ #include "pairwiseseqscommand.h" -#include "sequence.hpp" - -#include "gotohoverlap.hpp" -#include "needlemanoverlap.hpp" -#include "blastalign.hpp" -#include "noalign.hpp" -#include "nast.hpp" - -#include "ignoregaps.h" -#include "eachgapdist.h" -#include "eachgapignore.h" -#include "onegapdist.h" -#include "onegapignore.h" - //********************************************************************************************************************** -vector PairwiseSeqsCommand::getValidParameters(){ +vector PairwiseSeqsCommand::setParameters(){ try { - string AlignArray[] = {"fasta","align","match","mismatch","gapopen","gapextend", "processors","calc","compress","cutoff","countends","output","outputdir","inputdir"}; - vector myArray (AlignArray, AlignArray+(sizeof(AlignArray)/sizeof(string))); + CommandParameter pfasta("fasta", "InputTypes", "", "", "none", "none", "none","phylip-column",false,true,true); parameters.push_back(pfasta); + CommandParameter palign("align", "Multiple", "needleman-gotoh-blast-noalign", "needleman", "", "", "","",false,false); parameters.push_back(palign); + CommandParameter pmatch("match", "Number", "", "1.0", "", "", "","",false,false); parameters.push_back(pmatch); + CommandParameter pmismatch("mismatch", "Number", "", "-1.0", "", "", "","",false,false); parameters.push_back(pmismatch); + CommandParameter pgapopen("gapopen", "Number", "", "-2.0", "", "", "","",false,false); parameters.push_back(pgapopen); + CommandParameter pgapextend("gapextend", "Number", "", "-1.0", "", "", "","",false,false); parameters.push_back(pgapextend); + CommandParameter pprocessors("processors", "Number", "", "1", "", "", "","",false,false,true); parameters.push_back(pprocessors); + CommandParameter poutput("output", "Multiple", "column-lt-square-phylip", "column", "", "", "","phylip-column",false,false,true); parameters.push_back(poutput); + CommandParameter pcalc("calc", "Multiple", "nogaps-eachgap-onegap", "onegap", "", "", "","",false,false); parameters.push_back(pcalc); + CommandParameter pcountends("countends", "Boolean", "", "T", "", "", "","",false,false); parameters.push_back(pcountends); + CommandParameter pcompress("compress", "Boolean", "", "F", "", "", "","",false,false); parameters.push_back(pcompress); + CommandParameter pcutoff("cutoff", "Number", "", "1.0", "", "", "","",false,false,true); parameters.push_back(pcutoff); + CommandParameter pinputdir("inputdir", "String", "", "", "", "", "","",false,false); parameters.push_back(pinputdir); + CommandParameter poutputdir("outputdir", "String", "", "", "", "", "","",false,false); parameters.push_back(poutputdir); + + vector myArray; + for (int i = 0; i < parameters.size(); i++) { myArray.push_back(parameters[i].name); } return myArray; } catch(exception& e) { - m->errorOut(e, "PairwiseSeqsCommand", "getValidParameters"); + m->errorOut(e, "PairwiseSeqsCommand", "setParameters"); exit(1); } } //********************************************************************************************************************** -vector PairwiseSeqsCommand::getRequiredParameters(){ +string PairwiseSeqsCommand::getHelpString(){ try { - string AlignArray[] = {"fasta"}; - vector myArray (AlignArray, AlignArray+(sizeof(AlignArray)/sizeof(string))); - return myArray; + string helpString = ""; + helpString += "The pairwise.seqs command reads a fasta file and creates distance matrix.\n"; + helpString += "The pairwise.seqs command parameters are fasta, align, match, mismatch, gapopen, gapextend, calc, output, cutoff and processors.\n"; + helpString += "The fasta parameter is required. You may enter multiple fasta files by separating their names with dashes. ie. fasta=abrecovery.fasta-amzon.fasta \n"; + helpString += "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"; + helpString += "The match parameter allows you to specify the bonus for having the same base. The default is 1.0.\n"; + helpString += "The mistmatch parameter allows you to specify the penalty for having different bases. The default is -1.0.\n"; + helpString += "The gapopen parameter allows you to specify the penalty for opening a gap in an alignment. The default is -2.0.\n"; + helpString += "The gapextend parameter allows you to specify the penalty for extending a gap in an alignment. The default is -1.0.\n"; + helpString += "The calc parameter allows you to specify the method of calculating the distances. Your options are: nogaps, onegap or eachgap. The default is onegap.\n"; + helpString += "The countends parameter allows you to specify whether to include terminal gaps in distance. Your options are: T or F. The default is T.\n"; + helpString += "The cutoff parameter allows you to specify maximum distance to keep. The default is 1.0.\n"; + helpString += "The output parameter allows you to specify format of your distance matrix. Options are column, lt, and square. The default is column.\n"; + helpString += "The compress parameter allows you to indicate that you want the resulting distance file compressed. The default is false.\n"; + helpString += "The pairwise.seqs command should be in the following format: \n"; + helpString += "pairwise.seqs(fasta=yourfastaFile, align=yourAlignmentMethod) \n"; + helpString += "Example pairwise.seqs(fasta=candidate.fasta, align=blast)\n"; + helpString += "Note: No spaces between parameter labels (i.e. fasta), '=' and parameters (i.e.yourFastaFile).\n"; + return helpString; } catch(exception& e) { - m->errorOut(e, "PairwiseSeqsCommand", "getRequiredParameters"); + m->errorOut(e, "PairwiseSeqsCommand", "getHelpString"); exit(1); } } //********************************************************************************************************************** -vector PairwiseSeqsCommand::getRequiredFiles(){ - try { - vector myArray; - return myArray; - } - catch(exception& e) { - m->errorOut(e, "PairwiseSeqsCommand", "getRequiredFiles"); - exit(1); - } +string PairwiseSeqsCommand::getOutputPattern(string type) { + try { + string pattern = ""; + + if (type == "phylip") { pattern = "[filename],[outputtag],dist"; } + else if (type == "column") { pattern = "[filename],dist"; } + else { m->mothurOut("[ERROR]: No definition for type " + type + " output pattern.\n"); m->control_pressed = true; } + + return pattern; + } + catch(exception& e) { + m->errorOut(e, "PairwiseSeqsCommand", "getOutputPattern"); + exit(1); + } } //********************************************************************************************************************** PairwiseSeqsCommand::PairwiseSeqsCommand(){ try { - //initialize outputTypes + abort = true; calledHelp = true; + setParameters(); vector tempOutNames; outputTypes["phylip"] = tempOutNames; outputTypes["column"] = tempOutNames; @@ -74,16 +97,14 @@ PairwiseSeqsCommand::PairwiseSeqsCommand(){ //********************************************************************************************************************** PairwiseSeqsCommand::PairwiseSeqsCommand(string option) { try { - abort = false; + abort = false; calledHelp = false; //allow user to run help - if(option == "help") { help(); abort = true; } + if(option == "help") { help(); abort = true; calledHelp = true; } + else if(option == "citation") { citation(); abort = true; calledHelp = true;} else { - - //valid paramters for this command - string AlignArray[] = {"fasta","align","match","mismatch","gapopen","gapextend", "processors","cutoff","compress","calc","countends","output","outputdir","inputdir"}; - vector myArray (AlignArray, AlignArray+(sizeof(AlignArray)/sizeof(string))); + vector myArray = setParameters(); OptionParser parser(option); map parameters = parser.getParameters(); @@ -111,55 +132,77 @@ PairwiseSeqsCommand::PairwiseSeqsCommand(string option) { if (inputDir == "not found"){ inputDir = ""; } fastaFileName = validParameter.validFile(parameters, "fasta", false); - if (fastaFileName == "not found") { m->mothurOut("fasta is a required parameter for the pairwise.seqs command."); m->mothurOutEndLine(); abort = true; } + if (fastaFileName == "not found") { + //if there is a current fasta file, use it + string filename = m->getFastaFile(); + if (filename != "") { fastaFileNames.push_back(filename); m->mothurOut("Using " + filename + " as input file for the fasta parameter."); m->mothurOutEndLine(); } + else { m->mothurOut("You have no current fastafile and the fasta parameter is required."); m->mothurOutEndLine(); abort = true; } + } else { m->splitAtDash(fastaFileName, fastaFileNames); //go through files and make sure they are good, if not, then disregard them for (int i = 0; i < fastaFileNames.size(); i++) { - if (inputDir != "") { - string path = m->hasPath(fastaFileNames[i]); - //if the user has not given a path then, add inputdir. else leave path alone. - if (path == "") { fastaFileNames[i] = inputDir + fastaFileNames[i]; } - } - - int ableToOpen; - ifstream in; - - ableToOpen = m->openInputFile(fastaFileNames[i], in, "noerror"); - - //if you can't open it, try default location - if (ableToOpen == 1) { - if (m->getDefaultPath() != "") { //default path is set - string tryPath = m->getDefaultPath() + m->getSimpleName(fastaFileNames[i]); - m->mothurOut("Unable to open " + fastaFileNames[i] + ". Trying default " + tryPath); m->mothurOutEndLine(); - ifstream in2; - ableToOpen = m->openInputFile(tryPath, in2, "noerror"); - in2.close(); - fastaFileNames[i] = tryPath; + bool ignore = false; + if (fastaFileNames[i] == "current") { + fastaFileNames[i] = m->getFastaFile(); + if (fastaFileNames[i] != "") { m->mothurOut("Using " + fastaFileNames[i] + " as input file for the fasta parameter where you had given current."); m->mothurOutEndLine(); } + else { + m->mothurOut("You have no current fastafile, ignoring current."); m->mothurOutEndLine(); ignore=true; + //erase from file list + fastaFileNames.erase(fastaFileNames.begin()+i); + i--; } } - //if you can't open it, try output location - if (ableToOpen == 1) { - if (m->getOutputDir() != "") { //default path is set - string tryPath = m->getOutputDir() + m->getSimpleName(fastaFileNames[i]); - m->mothurOut("Unable to open " + fastaFileNames[i] + ". Trying output directory " + tryPath); m->mothurOutEndLine(); - ifstream in2; - ableToOpen = m->openInputFile(tryPath, in2, "noerror"); - in2.close(); - fastaFileNames[i] = tryPath; + if (!ignore) { + + if (inputDir != "") { + string path = m->hasPath(fastaFileNames[i]); + //if the user has not given a path then, add inputdir. else leave path alone. + if (path == "") { fastaFileNames[i] = inputDir + fastaFileNames[i]; } } - } + + int ableToOpen; + ifstream in; + + ableToOpen = m->openInputFile(fastaFileNames[i], in, "noerror"); - in.close(); + //if you can't open it, try default location + if (ableToOpen == 1) { + if (m->getDefaultPath() != "") { //default path is set + string tryPath = m->getDefaultPath() + m->getSimpleName(fastaFileNames[i]); + m->mothurOut("Unable to open " + fastaFileNames[i] + ". Trying default " + tryPath); m->mothurOutEndLine(); + ifstream in2; + ableToOpen = m->openInputFile(tryPath, in2, "noerror"); + in2.close(); + fastaFileNames[i] = tryPath; + } + } + + //if you can't open it, try output location + if (ableToOpen == 1) { + if (m->getOutputDir() != "") { //default path is set + string tryPath = m->getOutputDir() + m->getSimpleName(fastaFileNames[i]); + m->mothurOut("Unable to open " + fastaFileNames[i] + ". Trying output directory " + tryPath); m->mothurOutEndLine(); + ifstream in2; + ableToOpen = m->openInputFile(tryPath, in2, "noerror"); + in2.close(); + fastaFileNames[i] = tryPath; + } + } + + in.close(); - if (ableToOpen == 1) { - m->mothurOut("Unable to open " + fastaFileNames[i] + ". It will be disregarded."); m->mothurOutEndLine(); - //erase from file list - fastaFileNames.erase(fastaFileNames.begin()+i); - i--; + if (ableToOpen == 1) { + m->mothurOut("Unable to open " + fastaFileNames[i] + ". It will be disregarded."); m->mothurOutEndLine(); + //erase from file list + fastaFileNames.erase(fastaFileNames.begin()+i); + i--; + }else { + m->setFastaFile(fastaFileNames[i]); + } } } @@ -171,22 +214,26 @@ PairwiseSeqsCommand::PairwiseSeqsCommand(string option) { // ...at some point should added some additional type checking... string temp; temp = validParameter.validFile(parameters, "match", false); if (temp == "not found"){ temp = "1.0"; } - convert(temp, match); + m->mothurConvert(temp, match); temp = validParameter.validFile(parameters, "mismatch", false); if (temp == "not found"){ temp = "-1.0"; } - convert(temp, misMatch); + m->mothurConvert(temp, misMatch); + if (misMatch > 0) { m->mothurOut("[ERROR]: mismatch must be negative.\n"); abort=true; } temp = validParameter.validFile(parameters, "gapopen", false); if (temp == "not found"){ temp = "-2.0"; } - convert(temp, gapOpen); + m->mothurConvert(temp, gapOpen); + if (gapOpen > 0) { m->mothurOut("[ERROR]: gapopen must be negative.\n"); abort=true; } temp = validParameter.validFile(parameters, "gapextend", false); if (temp == "not found"){ temp = "-1.0"; } - convert(temp, gapExtend); + m->mothurConvert(temp, gapExtend); + if (gapExtend > 0) { m->mothurOut("[ERROR]: gapextend must be negative.\n"); abort=true; } - temp = validParameter.validFile(parameters, "processors", false); if (temp == "not found"){ temp = "1"; } - convert(temp, processors); + temp = validParameter.validFile(parameters, "processors", false); if (temp == "not found"){ temp = m->getProcessors(); } + m->setProcessors(temp); + m->mothurConvert(temp, processors); temp = validParameter.validFile(parameters, "cutoff", false); if(temp == "not found"){ temp = "1.0"; } - convert(temp, cutoff); + m->mothurConvert(temp, cutoff); temp = validParameter.validFile(parameters, "countends", false); if(temp == "not found"){ temp = "T"; } countends = m->isTrue(temp); @@ -197,6 +244,7 @@ PairwiseSeqsCommand::PairwiseSeqsCommand(string option) { align = validParameter.validFile(parameters, "align", false); if (align == "not found"){ align = "needleman"; } output = validParameter.validFile(parameters, "output", false); if(output == "not found"){ output = "column"; } + if (output=="phylip") { output = "lt"; } if ((output != "column") && (output != "lt") && (output != "square")) { m->mothurOut(output + " is not a valid output form. Options are column, lt and square. I will use column."); m->mothurOutEndLine(); output = "column"; } calc = validParameter.validFile(parameters, "calc", false); @@ -205,25 +253,6 @@ PairwiseSeqsCommand::PairwiseSeqsCommand(string option) { if (calc == "default") { calc = "onegap"; } } m->splitAtDash(calc, Estimators); - - ValidCalculators validCalculator; - if (countends) { - for (int i=0; imothurOut("The pairwise.seqs command reads a fasta file and creates distance matrix.\n"); - m->mothurOut("The pairwise.seqs command parameters are fasta, align, match, mismatch, gapopen, gapextend, calc, output, cutoff and processors.\n"); - m->mothurOut("The fasta parameter is required. You may enter multiple fasta files by separating their names with dashes. ie. fasta=abrecovery.fasta-amzon.fasta \n"); - m->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"); - m->mothurOut("The match parameter allows you to specify the bonus for having the same base. The default is 1.0.\n"); - m->mothurOut("The mistmatch parameter allows you to specify the penalty for having different bases. The default is -1.0.\n"); - m->mothurOut("The gapopen parameter allows you to specify the penalty for opening a gap in an alignment. The default is -2.0.\n"); - m->mothurOut("The gapextend parameter allows you to specify the penalty for extending a gap in an alignment. The default is -1.0.\n"); - m->mothurOut("The calc parameter allows you to specify the method of calculating the distances. Your options are: nogaps, onegap or eachgap. The default is onegap.\n"); - m->mothurOut("The countends parameter allows you to specify whether to include terminal gaps in distance. Your options are: T or F. The default is T.\n"); - m->mothurOut("The cutoff parameter allows you to specify maximum distance to keep. The default is 1.0.\n"); - m->mothurOut("The output parameter allows you to specify format of your distance matrix. Options are column, lt, and square. The default is column.\n"); - m->mothurOut("The compress parameter allows you to indicate that you want the resulting distance file compressed. The default is false.\n"); - m->mothurOut("The pairwise.seqs command should be in the following format: \n"); - m->mothurOut("pairwise.seqs(fasta=yourfastaFile, align=yourAlignmentMethod) \n"); - m->mothurOut("Example pairwise.seqs(fasta=candidate.fasta, align=blast)\n"); - m->mothurOut("Note: No spaces between parameter labels (i.e. fasta), '=' and parameters (i.e.yourFastaFile).\n\n"); - } - catch(exception& e) { - m->errorOut(e, "PairwiseSeqsCommand", "help"); - exit(1); - } -} - - //********************************************************************************************************************** int PairwiseSeqsCommand::execute(){ try { - if (abort == true) { return 0; } + if (abort == true) { if (calledHelp) { return 0; } return 2; } - int longestBase = 2000; //will need to update this in driver if we find sequences with more bases. hardcoded so we don't have the pre-read user fasta file. - - 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 { - m->mothurOut(align + " is not a valid alignment option. I will run the command using needleman."); - m->mothurOutEndLine(); - alignment = new NeedlemanOverlap(gapOpen, match, misMatch, longestBase); - } + longestBase = 2000; //will need to update this in driver if we find sequences with more bases. hardcoded so we don't have the pre-read user fasta file. cutoff += 0.005; @@ -298,17 +286,21 @@ int PairwiseSeqsCommand::execute(){ int numSeqs = alignDB.getNumSeqs(); int startTime = time(NULL); string outputFile = ""; - + + map variables; + variables["[filename]"] = outputDir + m->getRootName(m->getSimpleName(fastaFileNames[s])); if (output == "lt") { //does the user want lower triangle phylip formatted file - outputFile = outputDir + m->getRootName(m->getSimpleName(fastaFileNames[s])) + "phylip.dist"; - remove(outputFile.c_str()); outputTypes["phylip"].push_back(outputFile); + variables["[outputtag]"] = "phylip"; + outputFile = getOutputFileName("phylip", variables); + m->mothurRemove(outputFile); outputTypes["phylip"].push_back(outputFile); }else if (output == "column") { //user wants column format - outputFile = outputDir + m->getRootName(m->getSimpleName(fastaFileNames[s])) + "dist"; + outputFile = getOutputFileName("column", variables); outputTypes["column"].push_back(outputFile); - remove(outputFile.c_str()); + m->mothurRemove(outputFile); }else { //assume square - outputFile = outputDir + m->getRootName(m->getSimpleName(fastaFileNames[s])) + "square.dist"; - remove(outputFile.c_str()); + variables["[outputtag]"] = "square"; + outputFile = getOutputFileName("phylip", variables); + m->mothurRemove(outputFile); outputTypes["phylip"].push_back(outputFile); } @@ -346,11 +338,11 @@ int PairwiseSeqsCommand::execute(){ driverMPI(start, end, outMPI, cutoff); - if (m->control_pressed) { outputTypes.clear(); MPI_File_close(&outMPI); remove(outputFile.c_str()); delete distCalculator; return 0; } + if (m->control_pressed) { outputTypes.clear(); MPI_File_close(&outMPI); m->mothurRemove(outputFile); return 0; } //wait on chidren for(int i = 1; i < processors; i++) { - if (m->control_pressed) { outputTypes.clear(); MPI_File_close(&outMPI); remove(outputFile.c_str()); delete distCalculator; return 0; } + if (m->control_pressed) { outputTypes.clear(); MPI_File_close(&outMPI); m->mothurRemove(outputFile); return 0; } char buf[5]; MPI_Recv(buf, 5, MPI_CHAR, i, tag, MPI_COMM_WORLD, &status); @@ -359,7 +351,7 @@ int PairwiseSeqsCommand::execute(){ //do your part driverMPI(start, end, outMPI, cutoff); - if (m->control_pressed) { outputTypes.clear(); MPI_File_close(&outMPI); remove(outputFile.c_str()); delete distCalculator; return 0; } + if (m->control_pressed) { outputTypes.clear(); MPI_File_close(&outMPI); m->mothurRemove(outputFile); return 0; } char buf[5]; strcpy(buf, "done"); @@ -374,12 +366,12 @@ int PairwiseSeqsCommand::execute(){ //do your part string outputMyPart; - unsigned long int mySize; + unsigned long long mySize; if (output != "square"){ driverMPI(start, end, outputFile, mySize); } else { driverMPI(start, end, outputFile, mySize, output); } - if (m->control_pressed) { outputTypes.clear(); remove(outputFile.c_str()); delete distCalculator; return 0; } + if (m->control_pressed) { outputTypes.clear(); m->mothurRemove(outputFile); return 0; } int amode=MPI_MODE_APPEND|MPI_MODE_WRONLY|MPI_MODE_CREATE; // MPI_File outMPI; @@ -392,9 +384,9 @@ int PairwiseSeqsCommand::execute(){ //wait on chidren for(int b = 1; b < processors; b++) { - unsigned long int fileSize; + unsigned long long fileSize; - if (m->control_pressed) { outputTypes.clear(); MPI_File_close(&outMPI); remove(outputFile.c_str()); delete distCalculator; return 0; } + if (m->control_pressed) { outputTypes.clear(); MPI_File_close(&outMPI); m->mothurRemove(outputFile); return 0; } MPI_Recv(&fileSize, 1, MPI_LONG, b, tag, MPI_COMM_WORLD, &status); @@ -420,11 +412,11 @@ int PairwiseSeqsCommand::execute(){ MPI_File_close(&outMPI); }else { //you are a child process //do your part - unsigned long int size; + unsigned long long size; if (output != "square"){ driverMPI(start, end, (outputFile + toString(pid) + ".temp"), size); } else { driverMPI(start, end, (outputFile + toString(pid) + ".temp"), size, output); } - if (m->control_pressed) { delete distCalculator; return 0; } + if (m->control_pressed) { return 0; } //tell parent you are done. MPI_Send(&size, 1, MPI_LONG, 0, tag, MPI_COMM_WORLD); @@ -433,7 +425,7 @@ int PairwiseSeqsCommand::execute(){ MPI_Barrier(MPI_COMM_WORLD); //make everyone wait - just in case #else - #if defined (__APPLE__) || (__MACH__) || (linux) || (__linux) + //#if defined (__APPLE__) || (__MACH__) || (linux) || (__linux) || (__linux__) || (__unix__) || (__unix) //if you don't need to fork anything if(processors == 1){ if (output != "square") { driver(0, numSeqs, outputFile, cutoff); } @@ -441,36 +433,27 @@ int PairwiseSeqsCommand::execute(){ }else{ //you have multiple processors for (int i = 0; i < processors; i++) { - lines.push_back(new linePair()); + distlinePair tempLine; + lines.push_back(tempLine); if (output != "square") { - lines[i]->start = int (sqrt(float(i)/float(processors)) * numSeqs); - lines[i]->end = int (sqrt(float(i+1)/float(processors)) * numSeqs); + lines[i].start = int (sqrt(float(i)/float(processors)) * numSeqs); + lines[i].end = int (sqrt(float(i+1)/float(processors)) * numSeqs); }else{ - lines[i]->start = int ((float(i)/float(processors)) * numSeqs); - lines[i]->end = int ((float(i+1)/float(processors)) * numSeqs); + lines[i].start = int ((float(i)/float(processors)) * numSeqs); + lines[i].end = int ((float(i+1)/float(processors)) * numSeqs); } } - - createProcesses(outputFile); - - map::iterator it = processIDS.begin(); - rename((outputFile + toString(it->second) + ".temp").c_str(), outputFile.c_str()); - it++; - //append and remove temp files - for (; it != processIDS.end(); it++) { - m->appendFiles((outputFile + toString(it->second) + ".temp"), outputFile); - remove((outputFile + toString(it->second) + ".temp").c_str()); - } + createProcesses(outputFile); } - #else + //#else //ifstream inFASTA; - if (output != "square") { driver(0, numSeqs, outputFile, cutoff); } - else { driver(0, numSeqs, outputFile, "square"); } - #endif + //if (output != "square") { driver(0, numSeqs, outputFile, cutoff); } + //else { driver(0, numSeqs, outputFile, "square"); } + //#endif #endif - if (m->control_pressed) { outputTypes.clear(); delete distCalculator; remove(outputFile.c_str()); return 0; } + if (m->control_pressed) { outputTypes.clear(); m->mothurRemove(outputFile); return 0; } #ifdef USE_MPI MPI_Comm_rank(MPI_COMM_WORLD, &pid); @@ -498,13 +481,24 @@ int PairwiseSeqsCommand::execute(){ m->mothurOut("It took " + toString(time(NULL) - startTime) + " to calculate the distances for " + toString(numSeqs) + " sequences."); m->mothurOutEndLine(); - if (m->control_pressed) { outputTypes.clear(); delete distCalculator; remove(outputFile.c_str()); return 0; } + if (m->control_pressed) { outputTypes.clear(); m->mothurRemove(outputFile); return 0; } + } + + //set phylip file as new current phylipfile + string current = ""; + itTypes = outputTypes.find("phylip"); + if (itTypes != outputTypes.end()) { + if ((itTypes->second).size() != 0) { current = (itTypes->second)[0]; m->setPhylipFile(current); } + } + + //set column file as new current columnfile + itTypes = outputTypes.find("column"); + if (itTypes != outputTypes.end()) { + if ((itTypes->second).size() != 0) { current = (itTypes->second)[0]; m->setColumnFile(current); } } - - delete distCalculator; m->mothurOutEndLine(); - m->mothurOut("Output File Name: "); m->mothurOutEndLine(); + m->mothurOut("Output File Names: "); m->mothurOutEndLine(); for (int i = 0; i < outputNames.size(); i++) { m->mothurOut(outputNames[i]); m->mothurOutEndLine(); } m->mothurOutEndLine(); @@ -520,30 +514,88 @@ int PairwiseSeqsCommand::execute(){ /**************************************************************************************************/ void PairwiseSeqsCommand::createProcesses(string filename) { try { -#if defined (__APPLE__) || (__MACH__) || (linux) || (__linux) - int process = 0; + int process = 1; processIDS.clear(); + +#if defined (__APPLE__) || (__MACH__) || (linux) || (__linux) || (__linux__) || (__unix__) || (__unix) + //loop through and create all the processes you want while (process != processors) { int pid = fork(); if (pid > 0) { - processIDS[lines[process]->end] = pid; //create map from line number to pid so you can append files in correct order later + processIDS.push_back(pid); process++; }else if (pid == 0){ - if (output != "square") { driver(lines[process]->start, lines[process]->end, filename + toString(getpid()) + ".temp", cutoff); } - else { driver(lines[process]->start, lines[process]->end, filename + toString(getpid()) + ".temp", "square"); } + if (output != "square") { driver(lines[process].start, lines[process].end, filename + toString(getpid()) + ".temp", cutoff); } + else { driver(lines[process].start, lines[process].end, filename + toString(getpid()) + ".temp", "square"); } + exit(0); + }else { + m->mothurOut("[ERROR]: unable to spawn the necessary processes."); m->mothurOutEndLine(); + for (int i=0;imothurOut("unable to spawn the necessary processes."); m->mothurOutEndLine(); exit(0); } + } } - + + //parent do my part + if (output != "square") { driver(lines[0].start, lines[0].end, filename, cutoff); } + else { driver(lines[0].start, lines[0].end, filename, "square"); } + + //force parent to wait until all the processes are done - for (map::iterator it = processIDS.begin(); it != processIDS.end(); it++) { - int temp = it->second; + for (int i=0;i pDataArray; //[processors-1]; + DWORD dwThreadIdArray[processors-1]; + HANDLE hThreadArray[processors-1]; + + //Create processor-1 worker threads. + for( int i=0; icount != (pDataArray[i]->end-pDataArray[i]->start)) { + m->mothurOut("[ERROR]: process " + toString(i) + " only processed " + toString(pDataArray[i]->count) + " of " + toString(pDataArray[i]->end-pDataArray[i]->start) + " sequences assigned to it, quitting. \n"); m->control_pressed = true; + } + CloseHandle(hThreadArray[i]); + delete pDataArray[i]; + } + #endif + + //append and remove temp files + for (int i=0;iappendFiles((filename + toString(processIDS[i]) + ".temp"), filename); + m->mothurRemove((filename + toString(processIDS[i]) + ".temp")); + } + } catch(exception& e) { m->errorOut(e, "PairwiseSeqsCommand", "createProcesses"); @@ -557,6 +609,33 @@ int PairwiseSeqsCommand::driver(int startLine, int endLine, string dFileName, fl try { int startTime = time(NULL); + + Alignment* alignment; + 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 { + m->mothurOut(align + " is not a valid alignment option. I will run the command using needleman."); + m->mothurOutEndLine(); + alignment = new NeedlemanOverlap(gapOpen, match, misMatch, longestBase); + } + + ValidCalculators validCalculator; + Dist* distCalculator; + if (countends) { + if (validCalculator.isValidCalculator("distance", Estimators[0]) == true) { + if (Estimators[0] == "nogaps") { distCalculator = new ignoreGaps(); } + else if (Estimators[0] == "eachgap") { distCalculator = new eachGapDist(); } + else if (Estimators[0] == "onegap") { distCalculator = new oneGapDist(); } + } + }else { + if (validCalculator.isValidCalculator("distance", Estimators[0]) == true) { + if (Estimators[0] == "nogaps") { distCalculator = new ignoreGaps(); } + else if (Estimators[0] == "eachgap"){ distCalculator = new eachGapIgnoreTermGapDist(); } + else if (Estimators[0] == "onegap") { distCalculator = new oneGapIgnoreTermGapDist(); } + } + } //column file ofstream outFile(dFileName.c_str(), ios::trunc); @@ -576,7 +655,7 @@ int PairwiseSeqsCommand::driver(int startLine, int endLine, string dFileName, fl for(int j=0;jcontrol_pressed) { outFile.close(); return 0; } + if (m->control_pressed) { outFile.close(); delete alignment; delete distCalculator; return 0; } if (alignDB.get(i).getUnaligned().length() > alignment->getnRows()) { alignment->resize(alignDB.get(i).getUnaligned().length()+1); @@ -586,16 +665,21 @@ int PairwiseSeqsCommand::driver(int startLine, int endLine, string dFileName, fl alignment->resize(alignDB.get(j).getUnaligned().length()+1); } - Sequence* seqI = new Sequence(alignDB.get(i).getName(), alignDB.get(i).getAligned()); - Sequence* seqJ = new Sequence(alignDB.get(j).getName(), alignDB.get(j).getAligned()); + Sequence seqI(alignDB.get(i).getName(), alignDB.get(i).getAligned()); + Sequence seqJ(alignDB.get(j).getName(), alignDB.get(j).getAligned()); - Nast(alignment, seqI, seqJ); - - distCalculator->calcDist(*seqI, *seqJ); + alignment->align(seqI.getUnaligned(), seqJ.getUnaligned()); + seqI.setAligned(alignment->getSeqAAln()); + seqJ.setAligned(alignment->getSeqBAln()); + + //cout << seqI.getName() << '\t' << seqJ.getName() << endl; + //cout << alignment->getSeqAAln() << endl << alignment->getSeqBAln() << endl; + + distCalculator->calcDist(seqI, seqJ); double dist = distCalculator->getDist(); - - delete seqI; delete seqJ; - + + //cout << "dist = " << dist << endl; + if(dist <= cutoff){ if (output == "column") { outFile << alignDB.get(i).getName() << ' ' << alignDB.get(j).getName() << ' ' << dist << endl; } } @@ -612,6 +696,8 @@ int PairwiseSeqsCommand::driver(int startLine, int endLine, string dFileName, fl m->mothurOut(toString(endLine-1) + "\t" + toString(time(NULL) - startTime)); m->mothurOutEndLine(); outFile.close(); + delete alignment; + delete distCalculator; return 1; } @@ -626,7 +712,34 @@ int PairwiseSeqsCommand::driver(int startLine, int endLine, string dFileName, st try { int startTime = time(NULL); + + Alignment* alignment; + 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 { + m->mothurOut(align + " is not a valid alignment option. I will run the command using needleman."); + m->mothurOutEndLine(); + alignment = new NeedlemanOverlap(gapOpen, match, misMatch, longestBase); + } + ValidCalculators validCalculator; + Dist* distCalculator; + if (countends) { + if (validCalculator.isValidCalculator("distance", Estimators[0]) == true) { + if (Estimators[0] == "nogaps") { distCalculator = new ignoreGaps(); } + else if (Estimators[0] == "eachgap") { distCalculator = new eachGapDist(); } + else if (Estimators[0] == "onegap") { distCalculator = new oneGapDist(); } + } + }else { + if (validCalculator.isValidCalculator("distance", Estimators[0]) == true) { + if (Estimators[0] == "nogaps") { distCalculator = new ignoreGaps(); } + else if (Estimators[0] == "eachgap"){ distCalculator = new eachGapIgnoreTermGapDist(); } + else if (Estimators[0] == "onegap") { distCalculator = new oneGapIgnoreTermGapDist(); } + } + } + //column file ofstream outFile(dFileName.c_str(), ios::trunc); outFile.setf(ios::fixed, ios::showpoint); @@ -644,7 +757,7 @@ int PairwiseSeqsCommand::driver(int startLine, int endLine, string dFileName, st for(int j=0;jcontrol_pressed) { outFile.close(); return 0; } + if (m->control_pressed) { outFile.close(); delete alignment; delete distCalculator; return 0; } if (alignDB.get(i).getUnaligned().length() > alignment->getnRows()) { alignment->resize(alignDB.get(i).getUnaligned().length()+1); @@ -654,15 +767,15 @@ int PairwiseSeqsCommand::driver(int startLine, int endLine, string dFileName, st alignment->resize(alignDB.get(j).getUnaligned().length()+1); } - Sequence* seqI = new Sequence(alignDB.get(i).getName(), alignDB.get(i).getAligned()); - Sequence* seqJ = new Sequence(alignDB.get(j).getName(), alignDB.get(j).getAligned()); + Sequence seqI(alignDB.get(i).getName(), alignDB.get(i).getAligned()); + Sequence seqJ(alignDB.get(j).getName(), alignDB.get(j).getAligned()); - Nast(alignment, seqI, seqJ); + alignment->align(seqI.getUnaligned(), seqJ.getUnaligned()); + seqI.setAligned(alignment->getSeqAAln()); + seqJ.setAligned(alignment->getSeqBAln()); - distCalculator->calcDist(*seqI, *seqJ); + distCalculator->calcDist(seqI, seqJ); double dist = distCalculator->getDist(); - - delete seqI; delete seqJ; outFile << dist << '\t'; } @@ -677,6 +790,8 @@ int PairwiseSeqsCommand::driver(int startLine, int endLine, string dFileName, st m->mothurOut(toString(endLine-1) + "\t" + toString(time(NULL) - startTime)); m->mothurOutEndLine(); outFile.close(); + delete alignment; + delete distCalculator; return 1; } @@ -692,14 +807,41 @@ int PairwiseSeqsCommand::driverMPI(int startLine, int endLine, MPI_File& outMPI, try { MPI_Status status; int startTime = time(NULL); + + Alignment* alignment; + 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 { + m->mothurOut(align + " is not a valid alignment option. I will run the command using needleman."); + m->mothurOutEndLine(); + alignment = new NeedlemanOverlap(gapOpen, match, misMatch, longestBase); + } + ValidCalculators validCalculator; + Dist* distCalculator; + if (countends) { + if (validCalculator.isValidCalculator("distance", Estimators[0]) == true) { + if (Estimators[0] == "nogaps") { distCalculator = new ignoreGaps(); } + else if (Estimators[0] == "eachgap") { distCalculator = new eachGapDist(); } + else if (Estimators[0] == "onegap") { distCalculator = new oneGapDist(); } + } + }else { + if (validCalculator.isValidCalculator("distance", Estimators[0]) == true) { + if (Estimators[0] == "nogaps") { distCalculator = new ignoreGaps(); } + else if (Estimators[0] == "eachgap"){ distCalculator = new eachGapIgnoreTermGapDist(); } + else if (Estimators[0] == "onegap") { distCalculator = new oneGapIgnoreTermGapDist(); } + } + } + string outputString = ""; for(int i=startLine;icontrol_pressed) { return 0; } + if (m->control_pressed) { delete alignment; delete distCalculator; return 0; } if (alignDB.get(i).getUnaligned().length() > alignment->getnRows()) { alignment->resize(alignDB.get(i).getUnaligned().length()+1); @@ -709,16 +851,16 @@ int PairwiseSeqsCommand::driverMPI(int startLine, int endLine, MPI_File& outMPI, alignment->resize(alignDB.get(j).getUnaligned().length()+1); } - Sequence* seqI = new Sequence(alignDB.get(i).getName(), alignDB.get(i).getAligned()); - Sequence* seqJ = new Sequence(alignDB.get(j).getName(), alignDB.get(j).getAligned()); + Sequence seqI(alignDB.get(i).getName(), alignDB.get(i).getAligned()); + Sequence seqJ(alignDB.get(j).getName(), alignDB.get(j).getAligned()); - Nast(alignment, seqI, seqJ); + alignment->align(seqI.getUnaligned(), seqJ.getUnaligned()); + seqI.setAligned(alignment->getSeqAAln()); + seqJ.setAligned(alignment->getSeqBAln()); - distCalculator->calcDist(*seqI, *seqJ); + distCalculator->calcDist(seqI, seqJ); double dist = distCalculator->getDist(); - delete seqI; delete seqJ; - if(dist <= cutoff){ outputString += (alignDB.get(i).getName() + ' ' + alignDB.get(j).getName() + ' ' + toString(dist) + '\n'); } @@ -741,7 +883,8 @@ int PairwiseSeqsCommand::driverMPI(int startLine, int endLine, MPI_File& outMPI, delete buf; } - + delete alignment; + delete distCalculator; return 1; } catch(exception& e) { @@ -751,7 +894,7 @@ int PairwiseSeqsCommand::driverMPI(int startLine, int endLine, MPI_File& outMPI, } /**************************************************************************************************/ /////// need to fix to work with calcs and sequencedb -int PairwiseSeqsCommand::driverMPI(int startLine, int endLine, string file, unsigned long int& size){ +int PairwiseSeqsCommand::driverMPI(int startLine, int endLine, string file, unsigned long long& size){ try { MPI_Status status; @@ -763,7 +906,33 @@ int PairwiseSeqsCommand::driverMPI(int startLine, int endLine, string file, unsi MPI_File_open(MPI_COMM_SELF, filename, amode, MPI_INFO_NULL, &outMPI); - int startTime = time(NULL); + Alignment* alignment; + 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 { + m->mothurOut(align + " is not a valid alignment option. I will run the command using needleman."); + m->mothurOutEndLine(); + alignment = new NeedlemanOverlap(gapOpen, match, misMatch, longestBase); + } + + ValidCalculators validCalculator; + Dist* distCalculator; + if (countends) { + if (validCalculator.isValidCalculator("distance", Estimators[0]) == true) { + if (Estimators[0] == "nogaps") { distCalculator = new ignoreGaps(); } + else if (Estimators[0] == "eachgap") { distCalculator = new eachGapDist(); } + else if (Estimators[0] == "onegap") { distCalculator = new oneGapDist(); } + } + }else { + if (validCalculator.isValidCalculator("distance", Estimators[0]) == true) { + if (Estimators[0] == "nogaps") { distCalculator = new ignoreGaps(); } + else if (Estimators[0] == "eachgap"){ distCalculator = new eachGapIgnoreTermGapDist(); } + else if (Estimators[0] == "onegap") { distCalculator = new oneGapIgnoreTermGapDist(); } + } + } + string outputString = ""; size = 0; @@ -780,7 +949,7 @@ int PairwiseSeqsCommand::driverMPI(int startLine, int endLine, string file, unsi for(int j=0;jcontrol_pressed) { return 0; } + if (m->control_pressed) { delete alignment; delete distCalculator; return 0; } if (alignDB.get(i).getUnaligned().length() > alignment->getnRows()) { alignment->resize(alignDB.get(i).getUnaligned().length()+1); @@ -790,16 +959,16 @@ int PairwiseSeqsCommand::driverMPI(int startLine, int endLine, string file, unsi alignment->resize(alignDB.get(j).getUnaligned().length()+1); } - Sequence* seqI = new Sequence(alignDB.get(i).getName(), alignDB.get(i).getAligned()); - Sequence* seqJ = new Sequence(alignDB.get(j).getName(), alignDB.get(j).getAligned()); + Sequence seqI(alignDB.get(i).getName(), alignDB.get(i).getAligned()); + Sequence seqJ(alignDB.get(j).getName(), alignDB.get(j).getAligned()); - Nast(alignment, seqI, seqJ); + alignment->align(seqI.getUnaligned(), seqJ.getUnaligned()); + seqI.setAligned(alignment->getSeqAAln()); + seqJ.setAligned(alignment->getSeqBAln()); - distCalculator->calcDist(*seqI, *seqJ); + distCalculator->calcDist(seqI, seqJ); double dist = distCalculator->getDist(); - delete seqI; delete seqJ; - outputString += toString(dist) + "\t"; } @@ -817,6 +986,8 @@ int PairwiseSeqsCommand::driverMPI(int startLine, int endLine, string file, unsi } MPI_File_close(&outMPI); + delete alignment; + delete distCalculator; return 1; } @@ -827,7 +998,7 @@ int PairwiseSeqsCommand::driverMPI(int startLine, int endLine, string file, unsi } /**************************************************************************************************/ /////// need to fix to work with calcs and sequencedb -int PairwiseSeqsCommand::driverMPI(int startLine, int endLine, string file, unsigned long int& size, string square){ +int PairwiseSeqsCommand::driverMPI(int startLine, int endLine, string file, unsigned long long& size, string square){ try { MPI_Status status; @@ -839,8 +1010,33 @@ int PairwiseSeqsCommand::driverMPI(int startLine, int endLine, string file, unsi MPI_File_open(MPI_COMM_SELF, filename, amode, MPI_INFO_NULL, &outMPI); - int startTime = time(NULL); + Alignment* alignment; + 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 { + m->mothurOut(align + " is not a valid alignment option. I will run the command using needleman."); + m->mothurOutEndLine(); + alignment = new NeedlemanOverlap(gapOpen, match, misMatch, longestBase); + } + ValidCalculators validCalculator; + Dist* distCalculator; + if (countends) { + if (validCalculator.isValidCalculator("distance", Estimators[0]) == true) { + if (Estimators[0] == "nogaps") { distCalculator = new ignoreGaps(); } + else if (Estimators[0] == "eachgap") { distCalculator = new eachGapDist(); } + else if (Estimators[0] == "onegap") { distCalculator = new oneGapDist(); } + } + }else { + if (validCalculator.isValidCalculator("distance", Estimators[0]) == true) { + if (Estimators[0] == "nogaps") { distCalculator = new ignoreGaps(); } + else if (Estimators[0] == "eachgap"){ distCalculator = new eachGapIgnoreTermGapDist(); } + else if (Estimators[0] == "onegap") { distCalculator = new oneGapIgnoreTermGapDist(); } + } + } + string outputString = ""; size = 0; @@ -856,7 +1052,7 @@ int PairwiseSeqsCommand::driverMPI(int startLine, int endLine, string file, unsi for(int j=0;jcontrol_pressed) { return 0; } + if (m->control_pressed) { delete alignment; return 0; } if (alignDB.get(i).getUnaligned().length() > alignment->getnRows()) { alignment->resize(alignDB.get(i).getUnaligned().length()+1); @@ -866,16 +1062,16 @@ int PairwiseSeqsCommand::driverMPI(int startLine, int endLine, string file, unsi alignment->resize(alignDB.get(j).getUnaligned().length()+1); } - Sequence* seqI = new Sequence(alignDB.get(i).getName(), alignDB.get(i).getAligned()); - Sequence* seqJ = new Sequence(alignDB.get(j).getName(), alignDB.get(j).getAligned()); + Sequence seqI(alignDB.get(i).getName(), alignDB.get(i).getAligned()); + Sequence seqJ(alignDB.get(j).getName(), alignDB.get(j).getAligned()); - Nast(alignment, seqI, seqJ); + alignment->align(seqI.getUnaligned(), seqJ.getUnaligned()); + seqI.setAligned(alignment->getSeqAAln()); + seqJ.setAligned(alignment->getSeqBAln()); - distCalculator->calcDist(*seqI, *seqJ); + distCalculator->calcDist(seqI, seqJ); double dist = distCalculator->getDist(); - delete seqI; delete seqJ; - outputString += toString(dist) + "\t"; } @@ -894,6 +1090,7 @@ int PairwiseSeqsCommand::driverMPI(int startLine, int endLine, string file, unsi MPI_File_close(&outMPI); + delete alignment; return 1; } catch(exception& e) {