X-Git-Url: https://git.donarmstrong.com/?a=blobdiff_plain;f=makecontigscommand.cpp;h=560b39cdd6adfaec6fab5bafc8e30175038c0b9c;hb=e339f9008daa7d37c9a9034829565620a6abe783;hp=697521c16abc44dbeee019590f66fb4d03032c64;hpb=15b252e317fde72ce0a5de8776cb88112e0ed2b6;p=mothur.git diff --git a/makecontigscommand.cpp b/makecontigscommand.cpp index 697521c..560b39c 100644 --- a/makecontigscommand.cpp +++ b/makecontigscommand.cpp @@ -13,7 +13,7 @@ vector MakeContigsCommand::setParameters(){ try { CommandParameter pfasta("ffastq", "InputTypes", "", "", "none", "none", "none",false,true); parameters.push_back(pfasta); CommandParameter prfasta("rfastq", "InputTypes", "", "", "none", "none", "none",false,true); parameters.push_back(prfasta); - CommandParameter palign("align", "Multiple", "needleman-gotoh-blast-noalign", "needleman", "", "", "",false,false); parameters.push_back(palign); + CommandParameter palign("align", "Multiple", "needleman-gotoh", "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); @@ -63,6 +63,7 @@ MakeContigsCommand::MakeContigsCommand(){ vector tempOutNames; outputTypes["fasta"] = tempOutNames; outputTypes["qfile"] = tempOutNames; + outputTypes["mismatch"] = tempOutNames; } catch(exception& e) { m->errorOut(e, "MakeContigsCommand", "MakeContigsCommand"); @@ -96,6 +97,7 @@ MakeContigsCommand::MakeContigsCommand(string option) { vector tempOutNames; outputTypes["fasta"] = tempOutNames; outputTypes["qfile"] = tempOutNames; + outputTypes["mismatch"] = tempOutNames; //if the user changes the input directory command factory will send this info to us in the output parameter @@ -155,7 +157,7 @@ MakeContigsCommand::MakeContigsCommand(string option) { m->mothurConvert(temp, processors); align = validParameter.validFile(parameters, "align", false); if (align == "not found"){ align = "needleman"; } - if ((align != "needleman") && (align != "blast") && (align != "gotoh") && (align != "noalign")) { m->mothurOut(align + " is not a valid alignment method. Options are needleman, blast, gotoh and noalign. I will use needleman."); m->mothurOutEndLine(); align = "needleman"; } + if ((align != "needleman") && (align != "gotoh")) { m->mothurOut(align + " is not a valid alignment method. Options are needleman or gotoh. I will use needleman."); m->mothurOutEndLine(); align = "needleman"; } } } @@ -173,12 +175,31 @@ int MakeContigsCommand::execute(){ //this function will create a forward and reverse, fasta and qual files for each processor. //files has an entry for each processor. files[i][0] = forwardFasta, files[i][1] = forwardQual, files[i][2] = reverseFasta, files[i][3] = reverseQual int numReads = 0; - m->mothurOut("Reading fastq data..."); cout.flush(); + int start = time(NULL); + longestBase = 1000; + m->mothurOut("Reading fastq data...\n"); vector< vector > files = readFastqFiles(numReads); m->mothurOut("Done.\n"); + + if (m->control_pressed) { return 0; } + string outFastaFile = outputDir + m->getRootName(m->getSimpleName(ffastqfile)) + "contigs.fasta"; + string outQualFile = outputDir + m->getRootName(m->getSimpleName(ffastqfile)) + "contigs.qual"; + string outMisMatchFile = outputDir + m->getRootName(m->getSimpleName(ffastqfile)) + "contigs.mismatches"; + outputNames.push_back(outFastaFile); outputTypes["fasta"].push_back(outFastaFile); + outputNames.push_back(outQualFile); outputTypes["qfile"].push_back(outQualFile); + outputNames.push_back(outMisMatchFile); outputTypes["mismatch"].push_back(outMisMatchFile); - + m->mothurOut("Making contigs...\n"); + createProcesses(files, outFastaFile, outQualFile, outMisMatchFile); + m->mothurOut("Done.\n"); + + //remove temp fasta and qual files + for (int i = 0; i < processors; i++) { for(int j = 0; j < files[i].size(); j++) { m->mothurRemove(files[i][j]); } } + + if (m->control_pressed) { for (int i = 0; i < outputNames.size(); i++) { m->mothurRemove(outputNames[i]); } return 0; } + + m->mothurOut("It took " + toString(time(NULL) - start) + " secs to process " + toString(numReads) + " sequences.\n"); string currentFasta = ""; itTypes = outputTypes.find("fasta"); @@ -207,6 +228,225 @@ int MakeContigsCommand::execute(){ } } //********************************************************************************************************************** +int MakeContigsCommand::createProcesses(vector< vector > files, string outputFasta, string outputQual, string outputMisMatches) { + try { + int num = 0; + vector processIDS; +#if defined (__APPLE__) || (__MACH__) || (linux) || (__linux) || (__linux__) || (__unix__) || (__unix) + int process = 0; + + //loop through and create all the processes you want + while (process != processors-1) { + 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){ + num = driver(files[process], outputFasta + toString(getpid()) + ".temp", outputQual + toString(getpid()) + ".temp", outputMisMatches + toString(getpid()) + ".temp"); + + //pass numSeqs to parent + ofstream out; + string tempFile = outputFasta + toString(getpid()) + ".num.temp"; + m->openOutputFile(tempFile, out); + out << num << endl; + out.close(); + + exit(0); + }else { + m->mothurOut("[ERROR]: unable to spawn the necessary processes."); m->mothurOutEndLine(); + for (int i = 0; i < processIDS.size(); i++) { kill (processIDS[i], SIGINT); } + exit(0); + } + } + + //do my part + num = driver(files[processors-1], outputFasta, outputQual, outputMisMatches); + + //force parent to wait until all the processes are done + for (int i=0;iopenInputFile(tempFile, in); + if (!in.eof()) { int tempNum = 0; in >> tempNum; num += tempNum; } + in.close(); m->mothurRemove(tempFile); + } + #else + + ////////////////////////////////////////////////////////////////////////////////////////////////////// + //Windows version shared memory, so be careful when passing variables through the contigsData struct. + //Above fork() will clone, so memory is separate, but that's not the case with windows, + ////////////////////////////////////////////////////////////////////////////////////////////////////// + + vector pDataArray; + DWORD dwThreadIdArray[processors-1]; + HANDLE hThreadArray[processors-1]; + + //Create processor worker threads. + for( int i=0; icount; + CloseHandle(hThreadArray[i]); + delete pDataArray[i]; + } + + #endif + + for (int i = 0; i < processIDS.size(); i++) { + m->appendFiles((outputFasta + toString(processIDS[i]) + ".temp"), outputFasta); + m->mothurRemove((outputFasta + toString(processIDS[i]) + ".temp")); + + m->appendFiles((outputQual + toString(processIDS[i]) + ".temp"), outputQual); + m->mothurRemove((outputQual + toString(processIDS[i]) + ".temp")); + + m->appendFiles((outputMisMatches + toString(processIDS[i]) + ".temp"), outputMisMatches); + m->mothurRemove((outputMisMatches + toString(processIDS[i]) + ".temp")); + } + + return num; + } + catch(exception& e) { + m->errorOut(e, "MakeContigsCommand", "createProcesses"); + exit(1); + } +} +//********************************************************************************************************************** +int MakeContigsCommand::driver(vector files, string outputFasta, string outputQual, string outputMisMatches){ + try { + + Alignment* alignment; + if(align == "gotoh") { alignment = new GotohOverlap(gapOpen, gapExtend, match, misMatch, longestBase); } + else if(align == "needleman") { alignment = new NeedlemanOverlap(gapOpen, match, misMatch, longestBase); } + + int num = 0; + string thisffastafile = files[0]; + string thisfqualfile = files[1]; + string thisrfastafile = files[2]; + string thisrqualfile = files[3]; + + if (m->debug) { m->mothurOut("[DEBUG]: ffasta = " + thisffastafile + ".\n[DEBUG]: fqual = " + thisfqualfile + ".\n[DEBUG]: rfasta = " + thisrfastafile + ".\n[DEBUG]: rqual = " + thisrqualfile + ".\n"); } + + ifstream inFFasta, inRFasta, inFQual, inRQual; + m->openInputFile(thisffastafile, inFFasta); + m->openInputFile(thisfqualfile, inFQual); + m->openInputFile(thisrfastafile, inRFasta); + m->openInputFile(thisrqualfile, inRQual); + + ofstream outFasta, outQual, outMisMatch; + m->openOutputFile(outputFasta, outFasta); + m->openOutputFile(outputQual, outQual); + m->openOutputFile(outputMisMatches, outMisMatch); + outMisMatch << "Name\tLength\tMisMatches\n"; + + while ((!inFQual.eof()) && (!inFFasta.eof()) && (!inRFasta.eof()) && (!inRQual.eof())) { + + if (m->control_pressed) { break; } + + //read seqs and quality info + Sequence fSeq(inFFasta); m->gobble(inFFasta); + Sequence rSeq(inRFasta); m->gobble(inRFasta); + QualityScores fQual(inFQual); m->gobble(inFQual); + QualityScores rQual(inRQual); m->gobble(inRQual); + + //flip the reverse reads + rSeq.reverseComplement(); + rQual.flipQScores(); + + //pairwise align + alignment->align(fSeq.getUnaligned(), rSeq.getUnaligned()); + map ABaseMap = alignment->getSeqAAlnBaseMap(); + map BBaseMap = alignment->getSeqBAlnBaseMap(); + fSeq.setAligned(alignment->getSeqAAln()); + rSeq.setAligned(alignment->getSeqBAln()); + int length = fSeq.getAligned().length(); + + //traverse alignments merging into one contiguous seq + string contig = ""; + vector contigScores; + int numMismatches = 0; + string seq1 = fSeq.getAligned(); + string seq2 = rSeq.getAligned(); + + vector scores1 = fQual.getQualityScores(); + vector scores2 = rQual.getQualityScores(); + + for (int i = 0; i < length; i++) { + if (seq1[i] == seq2[i]) { //match, add base and choose highest score + contig += seq1[i]; + contigScores.push_back(scores1[ABaseMap[i]]); + if (scores1[ABaseMap[i]] < scores2[BBaseMap[i]]) { contigScores[i] = scores2[BBaseMap[i]]; } + }else if (((seq1[i] == '.') || (seq1[i] == '-')) && ((seq2[i] != '-') && (seq2[i] != '.'))) { //seq1 is a gap and seq2 is a base, choose seq2 + contig += seq2[i]; + contigScores.push_back(scores2[BBaseMap[i]]); + }else if (((seq2[i] == '.') || (seq2[i] == '-')) && ((seq1[i] != '-') && (seq1[i] != '.'))) { //seq2 is a gap and seq1 is a base, choose seq1 + contig += seq1[i]; + contigScores.push_back(scores1[ABaseMap[i]]); + }else if (((seq1[i] != '-') && (seq1[i] != '.')) && ((seq2[i] != '-') && (seq2[i] != '.'))) { //both bases choose one with better quality + char c = seq1[i]; + contigScores.push_back(scores1[ABaseMap[i]]); + if (scores1[ABaseMap[i]] < scores2[BBaseMap[i]]) { contigScores[i] = scores2[BBaseMap[i]]; c = seq2[i]; } + contig += c; + numMismatches++; + }else { //should never get here + m->mothurOut("[ERROR]: case I didn't think of seq1 = " + toString(seq1[i]) + " and seq2 = " + toString(seq2[i]) + "\n"); + } + } + + //output + outFasta << ">" << fSeq.getName() << endl << contig << endl; + outQual << ">" << fSeq.getName() << endl; + for (int i = 0; i < contigScores.size(); i++) { outQual << contigScores[i] << ' '; } + outQual << endl; + outMisMatch << fSeq.getName() << '\t' << contig.length() << '\t' << numMismatches << endl; + + num++; + + //report progress + if((num) % 1000 == 0){ m->mothurOut(toString(num)); m->mothurOutEndLine(); } + } + + //report progress + if((num) % 1000 != 0){ m->mothurOut(toString(num)); m->mothurOutEndLine(); } + + inFFasta.close(); + inFQual.close(); + inRFasta.close(); + inRQual.close(); + outFasta.close(); + outQual.close(); + outMisMatch.close(); + delete alignment; + + if (m->control_pressed) { m->mothurRemove(outputQual); m->mothurRemove(outputFasta); m->mothurRemove(outputMisMatches);} + + return num; + } + catch(exception& e) { + m->errorOut(e, "MakeContigsCommand", "driver"); + exit(1); + } +} +//********************************************************************************************************************** vector< vector > MakeContigsCommand::readFastqFiles(int& count){ try { vector< vector > files; @@ -287,7 +527,7 @@ vector< vector > MakeContigsCommand::readFastqFiles(int& count){ //report progress if((count) % 10000 != 0){ m->mothurOut(toString(count)); m->mothurOutEndLine(); } - + //close files, delete ofstreams for (it = tempfiles.begin(); it!=tempfiles.end(); it++) { for (int i = 0; i < (it->second).size(); i++) { (*(it->second)[i]).close(); delete (it->second)[i]; } } @@ -308,7 +548,6 @@ vector< vector > MakeContigsCommand::readFastqFiles(int& count){ exit(1); } } - //********************************************************************************************************************** fastqRead MakeContigsCommand::readFastq(ifstream& in){ try {