X-Git-Url: https://git.donarmstrong.com/?a=blobdiff_plain;f=prcseqscommand.cpp;h=6a4df6388683b1cda6aa5aa41d1ec3d2d672a15f;hb=a65438a717aba8cad1ff86c53279b191710b8810;hp=f2ca9db29d7a94ce0d22606d4ad114e11001fd4f;hpb=af0a94ea8f02421b2b73e77e68753a2b4c37768e;p=mothur.git diff --git a/prcseqscommand.cpp b/prcseqscommand.cpp index f2ca9db..6a4df63 100644 --- a/prcseqscommand.cpp +++ b/prcseqscommand.cpp @@ -312,6 +312,7 @@ int PcrSeqsCommand::execute(){ if (abort == true) { if (calledHelp) { return 0; } return 2; } int start = time(NULL); + fileAligned = true; string thisOutputDir = outputDir; if (outputDir == "") { thisOutputDir += m->hasPath(fastafile); } @@ -352,8 +353,7 @@ int PcrSeqsCommand::execute(){ if (m->control_pressed) { return 0; } set badNames; - if(processors == 1) { numFastaSeqs = driverPcr(fastafile, trimSeqFile, badSeqFile, badNames, lines[0]); } - else { numFastaSeqs = createProcesses(fastafile, trimSeqFile, badSeqFile, badNames); } + numFastaSeqs = createProcesses(fastafile, trimSeqFile, badSeqFile, badNames); if (m->control_pressed) { return 0; } @@ -429,6 +429,8 @@ int PcrSeqsCommand::createProcesses(string filename, string goodFileName, string vector processIDS; int process = 1; int num = 0; + int pstart = -1; int pend = -1; + bool adjustNeeded = false; #if defined (__APPLE__) || (__MACH__) || (linux) || (__linux) || (__linux__) || (__unix__) || (__unix) @@ -440,12 +442,14 @@ int PcrSeqsCommand::createProcesses(string filename, string goodFileName, string 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 = driverPcr(filename, goodFileName + toString(getpid()) + ".temp", badFileName + toString(getpid()) + ".temp", badSeqNames, lines[process]); + string locationsFile = toString(getpid()) + ".temp"; + num = driverPcr(filename, goodFileName + toString(getpid()) + ".temp", badFileName + toString(getpid()) + ".temp", locationsFile, badSeqNames, lines[process], pstart, adjustNeeded); //pass numSeqs to parent ofstream out; string tempFile = filename + toString(getpid()) + ".num.temp"; m->openOutputFile(tempFile, out); + out << pstart << '\t' << adjustNeeded << endl; out << num << '\t' << badSeqNames.size() << endl; for (set::iterator it = badSeqNames.begin(); it != badSeqNames.end(); it++) { out << (*it) << endl; @@ -460,7 +464,8 @@ int PcrSeqsCommand::createProcesses(string filename, string goodFileName, string } } - num = driverPcr(filename, goodFileName, badFileName, badSeqNames, lines[0]); + string locationsFile = toString(getpid()) + ".temp"; + num = driverPcr(filename, goodFileName, badFileName, locationsFile, badSeqNames, lines[0], pstart, adjustNeeded); //force parent to wait until all the processes are done for (int i=0;iopenInputFile(tempFile, in); int numBadNames = 0; string name = ""; - if (!in.eof()) { int tempNum = 0; in >> tempNum >> numBadNames; num += tempNum; m->gobble(in); } + int tpstart = -1; bool tempAdjust = false; + + if (!in.eof()) { + in >> tpstart >> tempAdjust; m->gobble(in); + + if (tempAdjust) { adjustNeeded = true; } + if (tpstart != -1) { + if (tpstart != pstart) { adjustNeeded = true; } + if (tpstart < pstart) { pstart = tpstart; } //smallest start + } + int tempNum = 0; in >> tempNum >> numBadNames; num += tempNum; m->gobble(in); + } for (int j = 0; j < numBadNames; j++) { in >> name; m->gobble(in); badSeqNames.insert(name); @@ -485,6 +501,9 @@ int PcrSeqsCommand::createProcesses(string filename, string goodFileName, string m->appendFiles((badFileName + toString(processIDS[i]) + ".temp"), badFileName); m->mothurRemove((badFileName + toString(processIDS[i]) + ".temp")); + + m->appendFiles((toString(processIDS[i]) + ".temp"), locationsFile); + m->mothurRemove((toString(processIDS[i]) + ".temp")); } #else @@ -498,6 +517,11 @@ int PcrSeqsCommand::createProcesses(string filename, string goodFileName, string DWORD dwThreadIdArray[processors-1]; HANDLE hThreadArray[processors-1]; + string locationsFile = "locationsFile.txt"; + m->mothurRemove(locationsFile); + m->mothurRemove(goodFileName); + m->mothurRemove(badFileName); + //Create processor worker threads. for( int i=0; icount != pDataArray[i]->fend) { m->mothurOut("[ERROR]: process " + toString(i) + " only processed " + toString(pDataArray[i]->count) + " of " + toString(pDataArray[i]->fend) + " sequences assigned to it, quitting. \n"); m->control_pressed = true; } + if (pDataArray[i]->adjustNeeded) { adjustNeeded = true; } + if (pDataArray[i]->pstart != -1) { + if (pDataArray[i]->pstart != pstart) { adjustNeeded = true; } + if (pDataArray[i]->pstart < pstart) { pstart = pDataArray[i]->pstart; } + } //smallest start + for (set::iterator it = pDataArray[i]->badSeqNames.begin(); it != pDataArray[i]->badSeqNames.end(); it++) { badSeqNames.insert(*it); } CloseHandle(hThreadArray[i]); delete pDataArray[i]; @@ -536,10 +566,50 @@ int PcrSeqsCommand::createProcesses(string filename, string goodFileName, string m->appendFiles((badFileName + toString(processIDS[i]) + ".temp"), badFileName); m->mothurRemove((badFileName + toString(processIDS[i]) + ".temp")); + + m->appendFiles((locationsFile+toString(processIDS[i]) + ".temp"), locationsFile); + m->mothurRemove((locationsFile+toString(processIDS[i]) + ".temp")); } #endif + + + if (fileAligned && adjustNeeded) { + //find pend - pend is the biggest ending value, but we must account for when we adjust the start. That adjustment may make the "new" end larger then the largest end. So lets find out what that "new" end will be. + ifstream inLocations; + m->openInputFile(locationsFile, inLocations); + + while(!inLocations.eof()) { + + if (m->control_pressed) { break; } + + string name = ""; + int thisStart = -1; int thisEnd = -1; + if (primers.size() != 0) { inLocations >> name >> thisStart; m->gobble(inLocations); } + if (revPrimer.size() != 0) { inLocations >> name >> thisEnd; m->gobble(inLocations); } + else { pend = -1; break; } + + int myDiff = 0; + if (pstart != -1) { + if (thisStart != -1) { + if (thisStart != pstart) { myDiff += (thisStart - pstart); } + } + } + + int myEnd = thisEnd + myDiff; + //cout << name << '\t' << thisStart << '\t' << thisEnd << " diff = " << myDiff << '\t' << myEnd << endl; + + if (thisEnd != -1) { + if (myEnd > pend) { pend = myEnd; } + } + + } + inLocations.close(); + + adjustDots(goodFileName, locationsFile, pstart, pend); + }else { m->mothurRemove(locationsFile); } + return num; } @@ -550,13 +620,16 @@ int PcrSeqsCommand::createProcesses(string filename, string goodFileName, string } //********************************************************************************************************************** -int PcrSeqsCommand::driverPcr(string filename, string goodFasta, string badFasta, set& badSeqNames, linePair filePos){ +int PcrSeqsCommand::driverPcr(string filename, string goodFasta, string badFasta, string locationsName, set& badSeqNames, linePair filePos, int& pstart, bool& adjustNeeded){ try { ofstream goodFile; m->openOutputFile(goodFasta, goodFile); ofstream badFile; m->openOutputFile(badFasta, badFile); + + ofstream locationsFile; + m->openOutputFile(locationsName, locationsFile); ifstream inFASTA; m->openInputFile(filename, inFASTA); @@ -566,6 +639,7 @@ int PcrSeqsCommand::driverPcr(string filename, string goodFasta, string badFasta bool done = false; int count = 0; set lengths; + set locations; //locations[0] = beginning locations, //pdiffs, bdiffs, primers, barcodes, revPrimers map faked; @@ -577,7 +651,15 @@ int PcrSeqsCommand::driverPcr(string filename, string goodFasta, string badFasta Sequence currSeq(inFASTA); m->gobble(inFASTA); + if (fileAligned) { //assume aligned until proven otherwise + lengths.insert(currSeq.getAligned().length()); + if (lengths.size() > 1) { fileAligned = false; } + } + string trashCode = ""; + string locationsString = ""; + int thisPStart = -1; + int thisPEnd = -1; if (currSeq.getName() != "") { if (m->debug) { m->mothurOut("[DEBUG]: seq name = " + currSeq.getName() + ".\n"); } @@ -596,13 +678,25 @@ int PcrSeqsCommand::driverPcr(string filename, string goodFasta, string badFasta else{ //are you aligned if (aligned) { - if (!keepprimer) { - if (keepdots) { currSeq.filterToPos(mapAligned[primerEnd]); } - else { currSeq.setAligned(currSeq.getAligned().substr(mapAligned[primerEnd])); } + if (!keepprimer) { + if (keepdots) { currSeq.filterToPos(mapAligned[primerEnd-1]+1); } //mapAligned[primerEnd-1] is the location of the last base in the primer. we want to trim to the space just after that. The -1 & +1 ensures if the primer is followed by gaps they are not trimmed causing an aligned sequence dataset to become unaligned. + else { + currSeq.setAligned(currSeq.getAligned().substr(mapAligned[primerEnd-1]+1)); + if (fileAligned) { + thisPStart = mapAligned[primerEnd-1]+1; //locations[0].insert(mapAligned[primerEnd-1]+1); + locationsString += currSeq.getName() + "\t" + toString(mapAligned[primerEnd-1]+1) + "\n"; + } + } } else { if (keepdots) { currSeq.filterToPos(mapAligned[primerStart]); } - else { currSeq.setAligned(currSeq.getAligned().substr(mapAligned[primerStart])); } + else { + currSeq.setAligned(currSeq.getAligned().substr(mapAligned[primerStart])); + if (fileAligned) { + thisPStart = mapAligned[primerStart]; //locations[0].insert(mapAligned[primerStart]); + locationsString += currSeq.getName() + "\t" + toString(mapAligned[primerStart]) + "\n"; + } + } } isAligned(currSeq.getAligned(), mapAligned); }else { @@ -617,16 +711,28 @@ int PcrSeqsCommand::driverPcr(string filename, string goodFasta, string badFasta int primerStart = 0; int primerEnd = 0; bool good = trim.findReverse(currSeq, primerStart, primerEnd); if(!good){ if (nomatch == "reject") { goodSeq = false; } trashCode += "r"; } - else{ - //are you aligned + else{ + //are you aligned if (aligned) { if (!keepprimer) { if (keepdots) { currSeq.filterFromPos(mapAligned[primerStart]); } - else { currSeq.setAligned(currSeq.getAligned().substr(0, mapAligned[primerStart])); } + else { + currSeq.setAligned(currSeq.getAligned().substr(0, mapAligned[primerStart])); + if (fileAligned) { + thisPEnd = mapAligned[primerStart]; //locations[1].insert(mapAligned[primerStart]); + locationsString += currSeq.getName() + "\t" + toString(mapAligned[primerStart]) + "\n"; + } + } } else { - if (keepdots) { currSeq.filterFromPos(mapAligned[primerEnd]); } - else { currSeq.setAligned(currSeq.getAligned().substr(0, mapAligned[primerEnd])); } + if (keepdots) { currSeq.filterFromPos(mapAligned[primerEnd-1]+1); } + else { + currSeq.setAligned(currSeq.getAligned().substr(0, mapAligned[primerEnd-1]+1)); + if (fileAligned) { + thisPEnd = mapAligned[primerEnd-1]+1; //locations[1].insert(mapAligned[primerEnd-1]+1); + locationsString += currSeq.getName() + "\t" + toString(mapAligned[primerEnd-1]+1) + "\n"; + } + } } } else { @@ -637,8 +743,7 @@ int PcrSeqsCommand::driverPcr(string filename, string goodFasta, string badFasta } }else if (ecolifile != "") { //make sure the seqs are aligned - lengths.insert(currSeq.getAligned().length()); - if (lengths.size() > 1) { m->mothurOut("[ERROR]: seqs are not aligned. When using start and end your sequences must be aligned.\n"); m->control_pressed = true; break; } + if (!fileAligned) { m->mothurOut("[ERROR]: seqs are not aligned. When using start and end your sequences must be aligned.\n"); m->control_pressed = true; break; } else if (currSeq.getAligned().length() != length) { m->mothurOut("[ERROR]: seqs are not the same length as ecoli seq. When using ecoli option your sequences must be aligned and the same length as the ecoli sequence.\n"); m->control_pressed = true; break; }else { @@ -653,8 +758,7 @@ int PcrSeqsCommand::driverPcr(string filename, string goodFasta, string badFasta } }else{ //using start and end to trim //make sure the seqs are aligned - lengths.insert(currSeq.getAligned().length()); - if (lengths.size() > 1) { m->mothurOut("[ERROR]: seqs are not aligned. When using start and end your sequences must be aligned.\n"); m->control_pressed = true; break; } + if (!fileAligned) { m->mothurOut("[ERROR]: seqs are not aligned. When using start and end your sequences must be aligned.\n"); m->control_pressed = true; break; } else { if (end != -1) { if (end > currSeq.getAligned().length()) { m->mothurOut("[ERROR]: end is longer than your sequence length, aborting.\n"); m->control_pressed = true; break; } @@ -679,7 +783,12 @@ int PcrSeqsCommand::driverPcr(string filename, string goodFasta, string badFasta //trimming removed all bases if (currSeq.getUnaligned() == "") { goodSeq = false; } - if(goodSeq == 1) { currSeq.printSequence(goodFile); } + if(goodSeq == 1) { + currSeq.printSequence(goodFile); + if (m->debug) { m->mothurOut("[DEBUG]: " + locationsString + "\n"); } + if (thisPStart != -1) { locations.insert(thisPStart); } + if (locationsString != "") { locationsFile << locationsString; } + } else { badSeqNames.insert(currSeq.getName()); currSeq.setName(currSeq.getName() + '|' + trashCode); @@ -704,7 +813,15 @@ int PcrSeqsCommand::driverPcr(string filename, string goodFasta, string badFasta badFile.close(); goodFile.close(); inFASTA.close(); - + locationsFile.close(); + + if (m->debug) { m->mothurOut("[DEBUG]: fileAligned = " + toString(fileAligned) +'\n'); } + + if (fileAligned && !keepdots) { //print out smallest start value and largest end value + if (locations.size() > 1) { adjustNeeded = true; } + if (primers.size() != 0) { set::iterator it = locations.begin(); pstart = *it; } + } + return count; } catch(exception& e) { @@ -731,6 +848,84 @@ bool PcrSeqsCommand::isAligned(string seq, map& aligned){ exit(1); } } +//********************************************************************************************************************** +int PcrSeqsCommand::adjustDots(string goodFasta, string locations, int pstart, int pend){ + try { + ifstream inFasta; + m->openInputFile(goodFasta, inFasta); + + ifstream inLocations; + m->openInputFile(locations, inLocations); + + ofstream out; + m->openOutputFile(goodFasta+".temp", out); + + set lengths; + //cout << pstart << '\t' << pend << endl; + //if (pstart > pend) { //swap them + + while(!inFasta.eof()) { + if(m->control_pressed) { break; } + + Sequence seq(inFasta); m->gobble(inFasta); + + string name = ""; + int thisStart = -1; int thisEnd = -1; + if (primers.size() != 0) { inLocations >> name >> thisStart; m->gobble(inLocations); } + if (revPrimer.size() != 0) { inLocations >> name >> thisEnd; m->gobble(inLocations); } + + + //cout << seq.getName() << '\t' << thisStart << '\t' << thisEnd << '\t' << seq.getAligned().length() << endl; + //cout << seq.getName() << '\t' << pstart << '\t' << pend << endl; + + if (name != seq.getName()) { m->mothurOut("[ERROR]: name mismatch in pcr.seqs.\n"); } + else { + if (pstart != -1) { + if (thisStart != -1) { + if (thisStart != pstart) { + string dots = ""; + for (int i = pstart; i < thisStart; i++) { dots += "."; } + thisEnd += dots.length(); + dots += seq.getAligned(); + seq.setAligned(dots); + } + } + } + + if (pend != -1) { + if (thisEnd != -1) { + if (thisEnd != pend) { + string dots = seq.getAligned(); + for (int i = thisEnd; i < pend; i++) { dots += "."; } + seq.setAligned(dots); + } + } + } + lengths.insert(seq.getAligned().length()); + } + + seq.printSequence(out); + } + inFasta.close(); + inLocations.close(); + out.close(); + m->mothurRemove(locations); + m->mothurRemove(goodFasta); + m->renameFile(goodFasta+".temp", goodFasta); + + //cout << "final lengths = \n"; + //for (set::iterator it = lengths.begin(); it != lengths.end(); it++) { + //cout << *it << endl; + // cout << lengths.count(*it) << endl; + // } + + return 0; + } + catch(exception& e) { + m->errorOut(e, "PcrSeqsCommand", "adjustDots"); + exit(1); + } +} //********************************************************************/ string PcrSeqsCommand::reverseOligo(string oligo){ try { @@ -808,14 +1003,35 @@ bool PcrSeqsCommand::readOligos(){ else if (c == 32 || c == 9){;} //space or tab } primers[oligo] = primerCount; primerCount++; + //cout << "for oligo = " << oligo << endl; }else if(type == "REVERSE"){ string oligoRC = reverseOligo(oligo); revPrimer.push_back(oligoRC); - //cout << "oligo = " << oligo << " reverse = " << oligoRC << endl; + //cout << "rev oligo = " << oligo << " reverse = " << oligoRC << endl; }else if(type == "BARCODE"){ - inOligos >> group; + inOligos >> group; + }else if(type == "PRIMER"){ + m->gobble(inOligos); + primers[oligo] = primerCount; primerCount++; + + string roligo=""; + inOligos >> roligo; + + for(int i=0;imothurOut(type + " is not recognized as a valid type. Choices are forward, reverse, linker, spacer and barcode. Ignoring " + oligo + "."); m->mothurOutEndLine(); m->control_pressed = true; } + else{ m->mothurOut(type + " is not recognized as a valid type. Choices are primer, forward, reverse, linker, spacer and barcode. Ignoring " + oligo + "."); m->mothurOutEndLine(); m->control_pressed = true; } } m->gobble(inOligos); } @@ -1088,7 +1304,7 @@ int PcrSeqsCommand::readCount(set badSeqNames){ //check for groups that have been eliminated CountTable ct; if (ct.testGroups(goodCountFile)) { - ct.readTable(goodCountFile, true); + ct.readTable(goodCountFile, true, false); ct.printTable(goodCountFile); }