X-Git-Url: https://git.donarmstrong.com/?a=blobdiff_plain;f=trimoligos.cpp;h=c35971a676f1c9af3de404bb5f0a4ca707a82f1f;hb=38029939675c5bf5e449eac90f801b35064b3a36;hp=245cfdcac8819d076782b85795300c0dfb034462;hpb=d6c0a11d1cecfac18b323285e7ffadb7f58e848f;p=mothur.git diff --git a/trimoligos.cpp b/trimoligos.cpp index 245cfdc..c35971a 100644 --- a/trimoligos.cpp +++ b/trimoligos.cpp @@ -28,6 +28,104 @@ TrimOligos::TrimOligos(int p, int b, int l, int s, map pr, map::iterator it=barcodes.begin();it!=barcodes.end();it++){ + if(it->first.length() > maxFBarcodeLength){ + maxFBarcodeLength = it->first.length(); + } + } + maxFPrimerLength = 0; + map::iterator it; + for(it=primers.begin();it!=primers.end();it++){ + if(it->first.length() > maxFPrimerLength){ + maxFPrimerLength = it->first.length(); + } + } + + maxLinkerLength = 0; + for(int i = 0; i < linker.size(); i++){ + if(linker[i].length() > maxLinkerLength){ + maxLinkerLength = linker[i].length(); + } + } + + maxSpacerLength = 0; + for(int i = 0; i < spacer.size(); i++){ + if(spacer[i].length() > maxSpacerLength){ + maxSpacerLength = spacer[i].length(); + } + } + } + catch(exception& e) { + m->errorOut(e, "TrimOligos", "TrimOligos"); + exit(1); + } +} +/********************************************************************/ +//strip, pdiffs, bdiffs, primers, barcodes, revPrimers +TrimOligos::TrimOligos(int p, int b, int l, int s, map pr, map br){ + try { + m = MothurOut::getInstance(); + + pdiffs = p; + bdiffs = b; + ldiffs = l; + sdiffs = s; + + maxFBarcodeLength = 0; + for(map::iterator it=br.begin();it!=br.end();it++){ + string forward = it->second.forward; + map >::iterator itForward = ifbarcodes.find(forward); + + if(forward.length() > maxFBarcodeLength){ maxFBarcodeLength = forward.length(); } + + if (itForward == ifbarcodes.end()) { + vector temp; temp.push_back(it->first); + ifbarcodes[forward] = temp; + }else { itForward->second.push_back(it->first); } + } + + maxFPrimerLength = 0; + for(map::iterator it=pr.begin();it!=pr.end();it++){ + string forward = it->second.forward; + map >::iterator itForward = ifprimers.find(forward); + + if(forward.length() > maxFPrimerLength){ maxFPrimerLength = forward.length(); } + + if (itForward == ifprimers.end()) { + vector temp; temp.push_back(it->first); + ifprimers[forward] = temp; + }else { itForward->second.push_back(it->first); } + } + + maxRBarcodeLength = 0; + for(map::iterator it=br.begin();it!=br.end();it++){ + string forward = it->second.reverse; + map >::iterator itForward = irbarcodes.find(forward); + + if(forward.length() > maxRBarcodeLength){ maxRBarcodeLength = forward.length(); } + + if (itForward == irbarcodes.end()) { + vector temp; temp.push_back(it->first); + irbarcodes[forward] = temp; + }else { itForward->second.push_back(it->first); } + } + + maxRPrimerLength = 0; + for(map::iterator it=pr.begin();it!=pr.end();it++){ + string forward = it->second.reverse; + map >::iterator itForward = irprimers.find(forward); + + if(forward.length() > maxRPrimerLength){ maxRPrimerLength = forward.length(); } + + if (itForward == irprimers.end()) { + vector temp; temp.push_back(it->first); + irprimers[forward] = temp; + }else { itForward->second.push_back(it->first); } + } + + ipbarcodes = br; + ipprimers = pr; } catch(exception& e) { m->errorOut(e, "TrimOligos", "TrimOligos"); @@ -86,21 +184,9 @@ int TrimOligos::stripBarcode(Sequence& seq, QualityScores& qual, int& group){ if ((bdiffs == 0) || (success == 0)) { return success; } else { //try aligning and see if you can find it - - int maxLength = 0; - Alignment* alignment; - if (barcodes.size() > 0) { - map::iterator it; - - for(it=barcodes.begin();it!=barcodes.end();it++){ - if(it->first.length() > maxLength){ - maxLength = it->first.length(); - } - } - alignment = new NeedlemanOverlap(-1.0, 1.0, -1.0, (maxLength+bdiffs+1)); - - }else{ alignment = NULL; } + if (barcodes.size() > 0) {alignment = new NeedlemanOverlap(-1.0, 1.0, -1.0, (maxFBarcodeLength+bdiffs+1)); } + else{ alignment = NULL; } //can you find the barcode int minDiff = 1e6; @@ -112,7 +198,7 @@ int TrimOligos::stripBarcode(Sequence& seq, QualityScores& qual, int& group){ string oligo = it->first; // int length = oligo.length(); - if(rawSequence.length() < maxLength){ //let's just assume that the barcodes are the same length + if(rawSequence.length() < maxFBarcodeLength){ //let's just assume that the barcodes are the same length success = bdiffs + 10; break; } @@ -129,8 +215,7 @@ int TrimOligos::stripBarcode(Sequence& seq, QualityScores& qual, int& group){ } oligo = oligo.substr(0,alnLength); temp = temp.substr(0,alnLength); - - int numDiff = countDiffs(oligo, temp); + int numDiff = countDiffs(oligo, temp); if(numDiff < minDiff){ minDiff = numDiff; @@ -154,7 +239,7 @@ int TrimOligos::stripBarcode(Sequence& seq, QualityScores& qual, int& group){ else{ //use the best match group = minGroup; seq.setUnaligned(rawSequence.substr(minPos)); - + if(qual.getName() != ""){ qual.trimQScores(minPos, -1); } @@ -172,6 +257,463 @@ int TrimOligos::stripBarcode(Sequence& seq, QualityScores& qual, int& group){ m->errorOut(e, "TrimOligos", "stripBarcode"); exit(1); } +} +//*******************************************************************/ +int TrimOligos::stripBarcode(Sequence& forwardSeq, Sequence& reverseSeq, QualityScores& forwardQual, QualityScores& reverseQual, int& group){ + try { + //look for forward barcode + string rawFSequence = forwardSeq.getUnaligned(); + string rawRSequence = reverseSeq.getUnaligned(); + int success = bdiffs + 1; //guilty until proven innocent + + //can you find the forward barcode + for(map::iterator it=ipbarcodes.begin();it!=ipbarcodes.end();it++){ + string foligo = it->second.forward; + string roligo = it->second.reverse; + + if(rawFSequence.length() < foligo.length()){ //let's just assume that the barcodes are the same length + success = bdiffs + 10; //if the sequence is shorter than the barcode then bail out + break; + } + if(rawRSequence.length() < roligo.length()){ //let's just assume that the barcodes are the same length + success = bdiffs + 10; //if the sequence is shorter than the barcode then bail out + break; + } + + if((compareDNASeq(foligo, rawFSequence.substr(0,foligo.length()))) && (compareDNASeq(roligo, rawRSequence.substr((rawRSequence.length()-roligo.length()),roligo.length())))) { + group = it->first; + forwardSeq.setUnaligned(rawFSequence.substr(foligo.length())); + reverseSeq.setUnaligned(rawRSequence.substr(0,(rawRSequence.length()-roligo.length()))); + forwardQual.trimQScores(foligo.length(), -1); + reverseQual.trimQScores(-1, rawRSequence.length()-roligo.length()); + success = 0; + break; + } + } + + //if you found the barcode or if you don't want to allow for diffs + if ((bdiffs == 0) || (success == 0)) { return success; } + else { //try aligning and see if you can find it + Alignment* alignment; + if (ifbarcodes.size() > 0) { alignment = new NeedlemanOverlap(-1.0, 1.0, -1.0, (maxFBarcodeLength+bdiffs+1)); } + else{ alignment = NULL; } + + //can you find the barcode + int minDiff = 1e6; + int minCount = 1; + vector< vector > minFGroup; + vector minFPos; + + //the pair can have duplicates, but we only want to search for the unique forward and reverses, example + /* + 1 Sarah Westcott + 2 John Westcott + 3 Anna Westcott + 4 Sarah Schloss + 5 Pat Schloss + 6 Gail Brown + 7 Pat Moore + + only want to look for forward = Sarah, John, Anna, Pat, Gail + reverse = Westcott, Schloss, Brown, Moore + + but if best match forward = 4, and reverse = 1, we want to count as a valid match because forward 1 and forward 4 are the same. so both barcodes map to same group. + */ + //cout << endl << forwardSeq.getName() << endl; + for(map >::iterator it=ifbarcodes.begin();it!=ifbarcodes.end();it++){ + string oligo = it->first; + + if(rawFSequence.length() < maxFBarcodeLength){ //let's just assume that the barcodes are the same length + success = bdiffs + 10; + break; + } + //cout << "before = " << oligo << '\t' << rawFSequence.substr(0,oligo.length()+bdiffs) << endl; + //use needleman to align first barcode.length()+numdiffs of sequence to each barcode + alignment->align(oligo, rawFSequence.substr(0,oligo.length()+bdiffs)); + oligo = alignment->getSeqAAln(); + string temp = alignment->getSeqBAln(); + + int alnLength = oligo.length(); + + for(int i=oligo.length()-1;i>=0;i--){ if(oligo[i] != '-'){ alnLength = i+1; break; } } + oligo = oligo.substr(0,alnLength); + temp = temp.substr(0,alnLength); + int numDiff = countDiffs(oligo, temp); + + if (alnLength == 0) { numDiff = bdiffs + 100; } + //cout << "after = " << oligo << '\t' << temp << '\t' << numDiff << endl; + + if(numDiff < minDiff){ + minDiff = numDiff; + minCount = 1; + minFGroup.clear(); + minFGroup.push_back(it->second); + int tempminFPos = 0; + minFPos.clear(); + for(int i=0;isecond); + int tempminFPos = 0; + for(int i=0;i bdiffs) { success = minDiff; } //no good matches + else{ + //check for reverse match + if (alignment != NULL) { delete alignment; } + + if (irbarcodes.size() > 0) { alignment = new NeedlemanOverlap(-1.0, 1.0, -1.0, (maxRBarcodeLength+bdiffs+1)); } + else{ alignment = NULL; } + + //can you find the barcode + minDiff = 1e6; + minCount = 1; + vector< vector > minRGroup; + vector minRPos; + + for(map >::iterator it=irbarcodes.begin();it!=irbarcodes.end();it++){ + string oligo = it->first; + //cout << "before = " << oligo << '\t' << rawRSequence.substr(0,oligo.length()+bdiffs) << endl; + if(rawRSequence.length() < maxRBarcodeLength){ //let's just assume that the barcodes are the same length + success = bdiffs + 10; + break; + } + + //use needleman to align first barcode.length()+numdiffs of sequence to each barcode + alignment->align(oligo, rawRSequence.substr(0,oligo.length()+bdiffs)); + oligo = alignment->getSeqAAln(); + string temp = alignment->getSeqBAln(); + + int alnLength = oligo.length(); + for(int i=oligo.length()-1;i>=0;i--){ if(oligo[i] != '-'){ alnLength = i+1; break; } } + oligo = oligo.substr(0,alnLength); + temp = temp.substr(0,alnLength); + int numDiff = countDiffs(oligo, temp); + if (alnLength == 0) { numDiff = bdiffs + 100; } + + //cout << "after = " << oligo << '\t' << temp << '\t' << numDiff << endl; + if(numDiff < minDiff){ + minDiff = numDiff; + minCount = 1; + minRGroup.clear(); + minRGroup.push_back(it->second); + int tempminRPos = 0; + minRPos.clear(); + for(int i=0;isecond); + } + + } + + /*cout << minDiff << '\t' << minCount << '\t' << endl; + for (int i = 0; i < minFGroup.size(); i++) { + cout << i << '\t'; + for (int j = 0; j < minFGroup[i].size(); j++) { cout << minFGroup[i][j] << " "; } + cout << endl; + } + cout << endl; + for (int i = 0; i < minRGroup.size(); i++) { + cout << i << '\t'; + for (int j = 0; j < minRGroup[i].size(); j++) { cout << minRGroup[i][j] << " "; } + cout << endl; + } + cout << endl;*/ + if(minDiff > bdiffs) { success = minDiff; } //no good matches + else { + bool foundMatch = false; + vector matches; + for (int i = 0; i < minFGroup.size(); i++) { + for (int j = 0; j < minFGroup[i].size(); j++) { + for (int k = 0; k < minRGroup.size(); k++) { + if (m->inUsersGroups(minFGroup[i][j], minRGroup[k])) { matches.push_back(minFGroup[i][j]); k+= minRGroup.size(); } + } + } + } + + int fStart = 0; + int rStart = 0; + if (matches.size() == 1) { + foundMatch = true; + group = matches[0]; + fStart = minFPos[0]; + rStart = minRPos[0]; + } + + //we have an acceptable match for the forward and reverse, but do they match? + if (foundMatch) { + forwardSeq.setUnaligned(rawFSequence.substr(fStart)); + reverseSeq.setUnaligned(rawRSequence.substr(rStart)); + forwardQual.trimQScores(fStart, -1); + reverseQual.trimQScores(rStart, -1); + success = minDiff; + }else { success = bdiffs + 100; } + } + } + + if (alignment != NULL) { delete alignment; } + } + + return success; + + } + catch(exception& e) { + m->errorOut(e, "TrimOligos", "stripIBarcode"); + exit(1); + } + +} +//*******************************************************************/ +int TrimOligos::stripForward(Sequence& forwardSeq, Sequence& reverseSeq, QualityScores& forwardQual, QualityScores& reverseQual, int& group){ + try { + //look for forward barcode + string rawFSequence = forwardSeq.getUnaligned(); + string rawRSequence = reverseSeq.getUnaligned(); + int success = pdiffs + 1; //guilty until proven innocent + + //can you find the forward barcode + for(map::iterator it=ipprimers.begin();it!=ipprimers.end();it++){ + string foligo = it->second.forward; + string roligo = it->second.reverse; + + if(rawFSequence.length() < foligo.length()){ //let's just assume that the barcodes are the same length + success = pdiffs + 10; //if the sequence is shorter than the barcode then bail out + break; + } + if(rawRSequence.length() < roligo.length()){ //let's just assume that the barcodes are the same length + success = pdiffs + 10; //if the sequence is shorter than the barcode then bail out + break; + } + + if((compareDNASeq(foligo, rawFSequence.substr(0,foligo.length()))) && (compareDNASeq(roligo, rawRSequence.substr((rawRSequence.length()-roligo.length()),roligo.length())))) { + group = it->first; + forwardSeq.setUnaligned(rawFSequence.substr(foligo.length())); + reverseSeq.setUnaligned(rawRSequence.substr(0,(rawRSequence.length()-roligo.length()))); + forwardQual.trimQScores(foligo.length(), -1); + reverseQual.trimQScores(-1, rawRSequence.length()-roligo.length()); + success = 0; + break; + } + } + + //if you found the barcode or if you don't want to allow for diffs + if ((pdiffs == 0) || (success == 0)) { return success; } + else { //try aligning and see if you can find it + Alignment* alignment; + if (ifprimers.size() > 0) { alignment = new NeedlemanOverlap(-1.0, 1.0, -1.0, (maxFPrimerLength+pdiffs+1)); } + else{ alignment = NULL; } + + //can you find the barcode + int minDiff = 1e6; + int minCount = 1; + vector< vector > minFGroup; + vector minFPos; + + //the pair can have duplicates, but we only want to search for the unique forward and reverses, example + /* + 1 Sarah Westcott + 2 John Westcott + 3 Anna Westcott + 4 Sarah Schloss + 5 Pat Schloss + 6 Gail Brown + 7 Pat Moore + + only want to look for forward = Sarah, John, Anna, Pat, Gail + reverse = Westcott, Schloss, Brown, Moore + + but if best match forward = 4, and reverse = 1, we want to count as a valid match because forward 1 and forward 4 are the same. so both barcodes map to same group. + */ + //cout << endl << forwardSeq.getName() << endl; + for(map >::iterator it=ifprimers.begin();it!=ifprimers.end();it++){ + string oligo = it->first; + + if(rawFSequence.length() < maxFPrimerLength){ //let's just assume that the barcodes are the same length + success = pdiffs + 10; + break; + } + //cout << "before = " << oligo << '\t' << rawFSequence.substr(0,oligo.length()+pdiffs) << endl; + //use needleman to align first barcode.length()+numdiffs of sequence to each barcode + alignment->align(oligo, rawFSequence.substr(0,oligo.length()+pdiffs)); + oligo = alignment->getSeqAAln(); + string temp = alignment->getSeqBAln(); + + int alnLength = oligo.length(); + + for(int i=oligo.length()-1;i>=0;i--){ if(oligo[i] != '-'){ alnLength = i+1; break; } } + oligo = oligo.substr(0,alnLength); + temp = temp.substr(0,alnLength); + int numDiff = countDiffs(oligo, temp); + + if (alnLength == 0) { numDiff = pdiffs + 100; } + //cout << "after = " << oligo << '\t' << temp << '\t' << numDiff << endl; + + if(numDiff < minDiff){ + minDiff = numDiff; + minCount = 1; + minFGroup.clear(); + minFGroup.push_back(it->second); + int tempminFPos = 0; + minFPos.clear(); + for(int i=0;isecond); + int tempminFPos = 0; + for(int i=0;i pdiffs) { success = minDiff; } //no good matches + else{ + //check for reverse match + if (alignment != NULL) { delete alignment; } + + if (irbarcodes.size() > 0) { alignment = new NeedlemanOverlap(-1.0, 1.0, -1.0, (maxRPrimerLength+pdiffs+1)); } + else{ alignment = NULL; } + + //can you find the barcode + minDiff = 1e6; + minCount = 1; + vector< vector > minRGroup; + vector minRPos; + + for(map >::iterator it=irprimers.begin();it!=irprimers.end();it++){ + string oligo = it->first; + //cout << "before = " << oligo << '\t' << rawRSequence.substr(0,oligo.length()+pdiffs) << endl; + if(rawRSequence.length() < maxRPrimerLength){ //let's just assume that the barcodes are the same length + success = pdiffs + 10; + break; + } + + //use needleman to align first barcode.length()+numdiffs of sequence to each barcode + alignment->align(oligo, rawRSequence.substr(0,oligo.length()+pdiffs)); + oligo = alignment->getSeqAAln(); + string temp = alignment->getSeqBAln(); + + int alnLength = oligo.length(); + for(int i=oligo.length()-1;i>=0;i--){ if(oligo[i] != '-'){ alnLength = i+1; break; } } + oligo = oligo.substr(0,alnLength); + temp = temp.substr(0,alnLength); + int numDiff = countDiffs(oligo, temp); + if (alnLength == 0) { numDiff = pdiffs + 100; } + + //cout << "after = " << oligo << '\t' << temp << '\t' << numDiff << endl; + if(numDiff < minDiff){ + minDiff = numDiff; + minCount = 1; + minRGroup.clear(); + minRGroup.push_back(it->second); + int tempminRPos = 0; + minRPos.clear(); + for(int i=0;isecond); + } + + } + + /*cout << minDiff << '\t' << minCount << '\t' << endl; + for (int i = 0; i < minFGroup.size(); i++) { + cout << i << '\t'; + for (int j = 0; j < minFGroup[i].size(); j++) { cout << minFGroup[i][j] << " "; } + cout << endl; + } + cout << endl; + for (int i = 0; i < minRGroup.size(); i++) { + cout << i << '\t'; + for (int j = 0; j < minRGroup[i].size(); j++) { cout << minRGroup[i][j] << " "; } + cout << endl; + } + cout << endl;*/ + if(minDiff > pdiffs) { success = minDiff; } //no good matches + else { + bool foundMatch = false; + vector matches; + for (int i = 0; i < minFGroup.size(); i++) { + for (int j = 0; j < minFGroup[i].size(); j++) { + for (int k = 0; k < minRGroup.size(); k++) { + if (m->inUsersGroups(minFGroup[i][j], minRGroup[k])) { matches.push_back(minFGroup[i][j]); k+= minRGroup.size(); } + } + } + } + + int fStart = 0; + int rStart = 0; + if (matches.size() == 1) { + foundMatch = true; + group = matches[0]; + fStart = minFPos[0]; + rStart = minRPos[0]; + } + + //we have an acceptable match for the forward and reverse, but do they match? + if (foundMatch) { + forwardSeq.setUnaligned(rawFSequence.substr(fStart)); + reverseSeq.setUnaligned(rawRSequence.substr(rStart)); + forwardQual.trimQScores(fStart, -1); + reverseQual.trimQScores(rStart, -1); + success = minDiff; + }else { success = pdiffs + 100; } + } + } + + if (alignment != NULL) { delete alignment; } + } + + return success; + + } + catch(exception& e) { + m->errorOut(e, "TrimOligos", "stripIForward"); + exit(1); + } } //*******************************************************************/ @@ -202,21 +744,9 @@ int TrimOligos::stripBarcode(Sequence& seq, int& group){ if ((bdiffs == 0) || (success == 0)) { return success; } else { //try aligning and see if you can find it - - int maxLength = 0; - - Alignment* alignment; - if (barcodes.size() > 0) { - map::iterator it=barcodes.begin(); - - for(map::iterator it=barcodes.begin();it!=barcodes.end();it++){ - if(it->first.length() > maxLength){ - maxLength = it->first.length(); - } - } - alignment = new NeedlemanOverlap(-1.0, 1.0, -1.0, (maxLength+bdiffs+1)); - - }else{ alignment = NULL; } + Alignment* alignment; + if (barcodes.size() > 0) { alignment = new NeedlemanOverlap(-1.0, 1.0, -1.0, (maxFBarcodeLength+bdiffs+1)); } + else{ alignment = NULL; } //can you find the barcode int minDiff = 1e6; @@ -228,7 +758,7 @@ int TrimOligos::stripBarcode(Sequence& seq, int& group){ string oligo = it->first; // int length = oligo.length(); - if(rawSequence.length() < maxLength){ //let's just assume that the barcodes are the same length + if(rawSequence.length() < maxFBarcodeLength){ //let's just assume that the barcodes are the same length success = bdiffs + 10; break; } @@ -237,7 +767,7 @@ int TrimOligos::stripBarcode(Sequence& seq, int& group){ alignment->align(oligo, rawSequence.substr(0,oligo.length()+bdiffs)); oligo = alignment->getSeqAAln(); string temp = alignment->getSeqBAln(); - + int alnLength = oligo.length(); for(int i=oligo.length()-1;i>=0;i--){ @@ -245,7 +775,7 @@ int TrimOligos::stripBarcode(Sequence& seq, int& group){ } oligo = oligo.substr(0,alnLength); temp = temp.substr(0,alnLength); - + int numDiff = countDiffs(oligo, temp); if(numDiff < minDiff){ @@ -286,7 +816,8 @@ int TrimOligos::stripBarcode(Sequence& seq, int& group){ } } -//********************************************************************/ + +/********************************************************************/ int TrimOligos::stripForward(Sequence& seq, int& group){ try { @@ -313,21 +844,9 @@ int TrimOligos::stripForward(Sequence& seq, int& group){ if ((pdiffs == 0) || (success == 0)) { return success; } else { //try aligning and see if you can find it - - int maxLength = 0; - Alignment* alignment; - if (primers.size() > 0) { - map::iterator it; - - for(it=primers.begin();it!=primers.end();it++){ - if(it->first.length() > maxLength){ - maxLength = it->first.length(); - } - } - alignment = new NeedlemanOverlap(-1.0, 1.0, -1.0, (maxLength+pdiffs+1)); - - }else{ alignment = NULL; } + if (primers.size() > 0) { alignment = new NeedlemanOverlap(-1.0, 1.0, -1.0, (maxFPrimerLength+pdiffs+1)); } + else{ alignment = NULL; } //can you find the barcode int minDiff = 1e6; @@ -339,7 +858,7 @@ int TrimOligos::stripForward(Sequence& seq, int& group){ string oligo = it->first; // int length = oligo.length(); - if(rawSequence.length() < maxLength){ + if(rawSequence.length() < maxFPrimerLength){ success = pdiffs + 100; break; } @@ -425,21 +944,9 @@ int TrimOligos::stripForward(Sequence& seq, QualityScores& qual, int& group, boo if ((pdiffs == 0) || (success == 0)) { return success; } else { //try aligning and see if you can find it - - int maxLength = 0; - Alignment* alignment; - if (primers.size() > 0) { - map::iterator it; - - for(it=primers.begin();it!=primers.end();it++){ - if(it->first.length() > maxLength){ - maxLength = it->first.length(); - } - } - alignment = new NeedlemanOverlap(-1.0, 1.0, -1.0, (maxLength+pdiffs+1)); - - }else{ alignment = NULL; } + if (primers.size() > 0) { alignment = new NeedlemanOverlap(-1.0, 1.0, -1.0, (maxFPrimerLength+pdiffs+1)); } + else{ alignment = NULL; } //can you find the barcode int minDiff = 1e6; @@ -451,7 +958,7 @@ int TrimOligos::stripForward(Sequence& seq, QualityScores& qual, int& group, boo string oligo = it->first; // int length = oligo.length(); - if(rawSequence.length() < maxLength){ + if(rawSequence.length() < maxFPrimerLength){ success = pdiffs + 100; break; } @@ -576,25 +1083,95 @@ bool TrimOligos::stripReverse(Sequence& seq){ bool TrimOligos::stripLinker(Sequence& seq, QualityScores& qual){ try { string rawSequence = seq.getUnaligned(); - bool success = 0; //guilty until proven innocent + bool success = ldiffs + 1; //guilty until proven innocent for(int i=0;i 0) { alignment = new NeedlemanOverlap(-1.0, 1.0, -1.0, (maxLinkerLength+ldiffs+1)); } + else{ alignment = NULL; } + + //can you find the barcode + int minDiff = 1e6; + int minCount = 1; + int minPos = 0; + + for(int i = 0; i < linker.size(); i++){ + string oligo = linker[i]; + // int length = oligo.length(); + + if(rawSequence.length() < maxLinkerLength){ //let's just assume that the barcodes are the same length + success = ldiffs + 10; + break; + } + + //use needleman to align first barcode.length()+numdiffs of sequence to each barcode + alignment->align(oligo, rawSequence.substr(0,oligo.length()+ldiffs)); + oligo = alignment->getSeqAAln(); + string temp = alignment->getSeqBAln(); + + int alnLength = oligo.length(); + + for(int i=oligo.length()-1;i>=0;i--){ + if(oligo[i] != '-'){ alnLength = i+1; break; } + } + oligo = oligo.substr(0,alnLength); + temp = temp.substr(0,alnLength); + + int numDiff = countDiffs(oligo, temp); + + if(numDiff < minDiff){ + minDiff = numDiff; + minCount = 1; + minPos = 0; + for(int i=0;i ldiffs) { success = minDiff; } //no good matches + else if(minCount > 1) { success = ldiffs + 100; } //can't tell the difference between multiple barcodes + else{ //use the best match + seq.setUnaligned(rawSequence.substr(minPos)); + + if(qual.getName() != ""){ + qual.trimQScores(minPos, -1); + } + success = minDiff; + } + + if (alignment != NULL) { delete alignment; } + + } + + return success; } @@ -608,23 +1185,87 @@ bool TrimOligos::stripLinker(Sequence& seq){ try { string rawSequence = seq.getUnaligned(); - bool success = 0; //guilty until proven innocent + bool success = ldiffs +1; //guilty until proven innocent for(int i=0;i 0) {alignment = new NeedlemanOverlap(-1.0, 1.0, -1.0, (maxLinkerLength+ldiffs+1)); } + else{ alignment = NULL; } + + //can you find the barcode + int minDiff = 1e6; + int minCount = 1; + int minPos = 0; + + for(int i = 0; i < linker.size(); i++){ + string oligo = linker[i]; + // int length = oligo.length(); + + if(rawSequence.length() < maxLinkerLength){ //let's just assume that the barcodes are the same length + success = ldiffs + 10; + break; + } + + //use needleman to align first barcode.length()+numdiffs of sequence to each barcode + alignment->align(oligo, rawSequence.substr(0,oligo.length()+ldiffs)); + oligo = alignment->getSeqAAln(); + string temp = alignment->getSeqBAln(); + + int alnLength = oligo.length(); + + for(int i=oligo.length()-1;i>=0;i--){ + if(oligo[i] != '-'){ alnLength = i+1; break; } + } + oligo = oligo.substr(0,alnLength); + temp = temp.substr(0,alnLength); + + int numDiff = countDiffs(oligo, temp); + + if(numDiff < minDiff){ + minDiff = numDiff; + minCount = 1; + minPos = 0; + for(int i=0;i ldiffs) { success = minDiff; } //no good matches + else if(minCount > 1) { success = ldiffs + 100; } //can't tell the difference between multiple barcodes + else{ //use the best match + seq.setUnaligned(rawSequence.substr(minPos)); + success = minDiff; + } + + if (alignment != NULL) { delete alignment; } + + } + return success; } @@ -638,25 +1279,95 @@ bool TrimOligos::stripLinker(Sequence& seq){ bool TrimOligos::stripSpacer(Sequence& seq, QualityScores& qual){ try { string rawSequence = seq.getUnaligned(); - bool success = 0; //guilty until proven innocent + bool success = sdiffs+1; //guilty until proven innocent for(int i=0;i 0) { alignment = new NeedlemanOverlap(-1.0, 1.0, -1.0, (maxSpacerLength+sdiffs+1)); } + else{ alignment = NULL; } + + //can you find the barcode + int minDiff = 1e6; + int minCount = 1; + int minPos = 0; + + for(int i = 0; i < spacer.size(); i++){ + string oligo = spacer[i]; + // int length = oligo.length(); + + if(rawSequence.length() < maxSpacerLength){ //let's just assume that the barcodes are the same length + success = sdiffs + 10; + break; + } + + //use needleman to align first barcode.length()+numdiffs of sequence to each barcode + alignment->align(oligo, rawSequence.substr(0,oligo.length()+sdiffs)); + oligo = alignment->getSeqAAln(); + string temp = alignment->getSeqBAln(); + + int alnLength = oligo.length(); + + for(int i=oligo.length()-1;i>=0;i--){ + if(oligo[i] != '-'){ alnLength = i+1; break; } + } + oligo = oligo.substr(0,alnLength); + temp = temp.substr(0,alnLength); + + int numDiff = countDiffs(oligo, temp); + + if(numDiff < minDiff){ + minDiff = numDiff; + minCount = 1; + minPos = 0; + for(int i=0;i sdiffs) { success = minDiff; } //no good matches + else if(minCount > 1) { success = sdiffs + 100; } //can't tell the difference between multiple barcodes + else{ //use the best match + seq.setUnaligned(rawSequence.substr(minPos)); + + if(qual.getName() != ""){ + qual.trimQScores(minPos, -1); + } + success = minDiff; + } + + if (alignment != NULL) { delete alignment; } + + } + + return success; } @@ -670,23 +1381,87 @@ bool TrimOligos::stripSpacer(Sequence& seq){ try { string rawSequence = seq.getUnaligned(); - bool success = 0; //guilty until proven innocent + bool success = sdiffs+1; //guilty until proven innocent for(int i=0;i 0) { alignment = new NeedlemanOverlap(-1.0, 1.0, -1.0, (maxSpacerLength+sdiffs+1)); } + else{ alignment = NULL; } + + //can you find the barcode + int minDiff = 1e6; + int minCount = 1; + int minPos = 0; + + for(int i = 0; i < spacer.size(); i++){ + string oligo = spacer[i]; + // int length = oligo.length(); + + if(rawSequence.length() < maxSpacerLength){ //let's just assume that the barcodes are the same length + success = sdiffs + 10; + break; + } + + //use needleman to align first barcode.length()+numdiffs of sequence to each barcode + alignment->align(oligo, rawSequence.substr(0,oligo.length()+sdiffs)); + oligo = alignment->getSeqAAln(); + string temp = alignment->getSeqBAln(); + + int alnLength = oligo.length(); + + for(int i=oligo.length()-1;i>=0;i--){ + if(oligo[i] != '-'){ alnLength = i+1; break; } + } + oligo = oligo.substr(0,alnLength); + temp = temp.substr(0,alnLength); + + int numDiff = countDiffs(oligo, temp); + + if(numDiff < minDiff){ + minDiff = numDiff; + minCount = 1; + minPos = 0; + for(int i=0;i sdiffs) { success = minDiff; } //no good matches + else if(minCount > 1) { success = sdiffs + 100; } //can't tell the difference between multiple barcodes + else{ //use the best match + seq.setUnaligned(rawSequence.substr(minPos)); + success = minDiff; + } + + if (alignment != NULL) { delete alignment; } + + } + return success; }