X-Git-Url: https://git.donarmstrong.com/?a=blobdiff_plain;f=trimoligos.cpp;h=91bdb83ad2d3a7f9e97713dba66709018d89908c;hb=e9845ee4c8db2e044e87d721cc2d94f8d609e03d;hp=8f4cbe98a7488b9d552f07b6ffd17aeb5e858c23;hpb=90708fe9701e3827e477c82fb3652539c3bf2a0d;p=mothur.git diff --git a/trimoligos.cpp b/trimoligos.cpp index 8f4cbe9..91bdb83 100644 --- a/trimoligos.cpp +++ b/trimoligos.cpp @@ -17,6 +17,8 @@ TrimOligos::TrimOligos(int p, int b, int l, int s, map pr, map br, vector r, vector lk, vector sp){ try { m = MothurOut::getInstance(); + paired = false; + trashCode = ""; pdiffs = p; bdiffs = b; @@ -28,6 +30,33 @@ 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"); @@ -36,7 +65,7 @@ TrimOligos::TrimOligos(int p, int b, int l, int s, map pr, map pr, map br, vector lk, vector sp){ +TrimOligos::TrimOligos(int p, int b, int l, int s, map pr, map br){ try { m = MothurOut::getInstance(); @@ -44,11 +73,63 @@ TrimOligos::TrimOligos(int p, int b, int l, int s, map pr, map< bdiffs = b; ldiffs = l; sdiffs = s; + paired = true; + trashCode = ""; - ibarcodes = br; - iprimers = pr; - linker = lk; - spacer = sp; + 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"); @@ -67,6 +148,26 @@ TrimOligos::TrimOligos(int p, int b, map pr, map br, v barcodes = br; primers = pr; revPrimer = r; + paired = false; + trashCode = ""; + + maxFBarcodeLength = 0; + for(map::iterator it=barcodes.begin();it!=barcodes.end();it++){ + string oligo = it->first; + if(oligo.length() > maxFBarcodeLength){ + maxFBarcodeLength = oligo.length(); + } + } + maxRBarcodeLength = maxFBarcodeLength; + + maxFPrimerLength = 0; + for(map::iterator it=primers.begin();it!=primers.end();it++){ + string oligo = it->first; + if(oligo.length() > maxFPrimerLength){ + maxFPrimerLength = oligo.length(); + } + } + maxRPrimerLength = maxFPrimerLength; } catch(exception& e) { m->errorOut(e, "TrimOligos", "TrimOligos"); @@ -75,10 +176,143 @@ TrimOligos::TrimOligos(int p, int b, map pr, map br, v } /********************************************************************/ TrimOligos::~TrimOligos() {} +//********************************************************************/ +bool TrimOligos::findForward(Sequence& seq, int& primerStart, int& primerEnd){ + try { + + string rawSequence = seq.getUnaligned(); + trashCode = ""; + + for(map::iterator it=primers.begin();it!=primers.end();it++){ + string oligo = it->first; + + if(rawSequence.length() < oligo.length()) { break; } + + //search for primer + int olength = oligo.length(); + for (int j = 0; j < rawSequence.length()-olength; j++){ + if (m->control_pressed) { primerStart = 0; primerEnd = 0; return false; } + string rawChunk = rawSequence.substr(j, olength); + if(compareDNASeq(oligo, rawChunk)) { + primerStart = j; + primerEnd = primerStart + olength; + return true; + } + + } + } + + primerStart = 0; primerEnd = 0; + //if you don't want to allow for diffs + if (pdiffs == 0) { trashCode = "f"; return false; } + else { //try aligning and see if you can find it + + //can you find the barcode + int minDiff = 1e6; + int minCount = 1; + + Alignment* alignment; + if (primers.size() > 0) { alignment = new NeedlemanOverlap(-1.0, 1.0, -1.0, (maxFPrimerLength+pdiffs+1)); } + else{ alignment = NULL; } + + for(map::iterator it=primers.begin();it!=primers.end();it++){ + + string prim = it->first; + //search for primer + int olength = prim.length(); + if (rawSequence.length() < olength) {} //ignore primers too long for this seq + else{ + for (int j = 0; j < rawSequence.length()-olength; j++){ + + string oligo = it->first; + + if (m->control_pressed) { primerStart = 0; primerEnd = 0; return false; } + + string rawChunk = rawSequence.substr(j, olength+pdiffs); + + //use needleman to align first primer.length()+numdiffs of sequence to each barcode + alignment->alignPrimer(oligo, rawChunk); + 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; + primerStart = j; + primerEnd = primerStart + alnLength; + }else if(numDiff == minDiff){ minCount++; } + } + } + } + + if (alignment != NULL) { delete alignment; } + + if(minDiff > pdiffs) { primerStart = 0; primerEnd = 0; trashCode = "f"; return false; } //no good matches + else if(minCount > 1) { primerStart = 0; primerEnd = 0; trashCode = "f"; return false; } //can't tell the difference between multiple primers + else{ trashCode = ""; return true; } + } + + trashCode = "f"; + primerStart = 0; primerEnd = 0; + return false; + + } + catch(exception& e) { + m->errorOut(e, "TrimOligos", "stripForward"); + exit(1); + } +} +//******************************************************************/ +bool TrimOligos::findReverse(Sequence& seq, int& primerStart, int& primerEnd){ + try { + trashCode = ""; + string rawSequence = seq.getUnaligned(); + + for(int i=0;i= 0; j--){ + if (m->control_pressed) { primerStart = 0; primerEnd = 0; return false; } + string rawChunk = rawSequence.substr(j, olength); + + if(compareDNASeq(oligo, rawChunk)) { + primerStart = j; + primerEnd = primerStart + olength; + return true; + } + + } + } + + trashCode = "r"; + primerStart = 0; primerEnd = 0; + return false; + } + catch(exception& e) { + m->errorOut(e, "PcrSeqsCommand", "findReverse"); + exit(1); + } +} //*******************************************************************/ + int TrimOligos::stripBarcode(Sequence& seq, QualityScores& qual, int& group){ try { - + trashCode = ""; + if (paired) { int success = stripPairedBarcode(seq, qual, group); return success; } + string rawSequence = seq.getUnaligned(); int success = bdiffs + 1; //guilty until proven innocent @@ -99,29 +333,18 @@ int TrimOligos::stripBarcode(Sequence& seq, QualityScores& qual, int& group){ } success = 0; - break; + trashCode = ""; + return success; } } //if you found the barcode or if you don't want to allow for diffs - if ((bdiffs == 0) || (success == 0)) { return success; } + if (bdiffs == 0) { trashCode = "b"; 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; @@ -133,13 +356,13 @@ 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 - success = bdiffs + 10; + if(rawSequence.length() < maxFBarcodeLength){ //let's just assume that the barcodes are the same length + success = bdiffs + 10; trashCode = "b"; break; } //use needleman to align first barcode.length()+numdiffs of sequence to each barcode - alignment->align(oligo, rawSequence.substr(0,oligo.length()+bdiffs)); + alignment->alignPrimer(oligo, rawSequence.substr(0,oligo.length()+bdiffs)); oligo = alignment->getSeqAAln(); string temp = alignment->getSeqBAln(); @@ -169,8 +392,8 @@ int TrimOligos::stripBarcode(Sequence& seq, QualityScores& qual, int& group){ } - if(minDiff > bdiffs) { success = minDiff; } //no good matches - else if(minCount > 1) { success = bdiffs + 100; } //can't tell the difference between multiple barcodes + if(minDiff > bdiffs) { trashCode = "b"; success = minDiff; } //no good matches + else if(minCount > 1) { trashCode = "b"; success = bdiffs + 100; } //can't tell the difference between multiple barcodes else{ //use the best match group = minGroup; seq.setUnaligned(rawSequence.substr(minPos)); @@ -179,6 +402,7 @@ int TrimOligos::stripBarcode(Sequence& seq, QualityScores& qual, int& group){ qual.trimQScores(minPos, -1); } success = minDiff; + trashCode = ""; } if (alignment != NULL) { delete alignment; } @@ -194,72 +418,87 @@ int TrimOligos::stripBarcode(Sequence& seq, QualityScores& qual, int& group){ } } //*******************************************************************/ -int TrimOligos::stripBarcode(Sequence& forwardSeq, Sequence& reverseSeq, QualityScores& forwardQual, QualityScores& reverseQual, int& group){ +int TrimOligos::stripBarcode(Sequence& forwardSeq, Sequence& reverseSeq, int& group){ try { + trashCode = ""; //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=ibarcodes.begin();it!=ibarcodes.end();it++){ + 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; + trashCode += "b"; + 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 + trashCode += "d"; 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(compareDNASeq(foligo, rawFSequence.substr(0,foligo.length()))) { + if (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()))); + success = 0; trashCode = ""; + return success; + } + }else { + trashCode = "b"; + if (!compareDNASeq(roligo, rawRSequence.substr((rawRSequence.length()-roligo.length()),roligo.length()))) { trashCode += "d"; } + } } //if you found the barcode or if you don't want to allow for diffs - if ((bdiffs == 0) || (success == 0)) { return success; } + if (bdiffs == 0) { return success; } else { //try aligning and see if you can find it - - //look for forward - int maxLength = 0; - + trashCode = ""; Alignment* alignment; - if (ibarcodes.size() > 0) { - for(map::iterator it=ibarcodes.begin();it!=ibarcodes.end();it++){ - if(it->second.forward.length() > maxLength){ maxLength = it->second.forward.length(); } - } - alignment = new NeedlemanOverlap(-1.0, 1.0, -1.0, (maxLength+bdiffs+1)); - }else{ alignment = NULL; } + 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; - int minFGroup = -1; - int minFPos = 0; - - for(map::iterator it=ibarcodes.begin();it!=ibarcodes.end();it++){ - string oligo = it->second.forward; + 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() < maxLength){ //let's just assume that the barcodes are the same length + 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)); + alignment->alignPrimer(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; } } @@ -267,95 +506,145 @@ int TrimOligos::stripBarcode(Sequence& forwardSeq, Sequence& reverseSeq, Quality 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 = it->first; - minFPos = 0; + 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 if(minCount > 1) { success = bdiffs + 100; } //can't tell the difference between multiple barcodes - else{ + + //cout << minDiff << '\t' << minCount << '\t' << endl; + if(minDiff > bdiffs) { trashCode = "b"; minFGroup.clear(); success = minDiff; } //no good matches + //else{ //check for reverse match if (alignment != NULL) { delete alignment; } - maxLength = 0; - if (ibarcodes.size() > 0) { - for(map::iterator it=ibarcodes.begin();it!=ibarcodes.end();it++){ - if(it->second.reverse.length() > maxLength){ maxLength = it->second.reverse.length(); } - } - alignment = new NeedlemanOverlap(-1.0, 1.0, -1.0, (maxLength+bdiffs+1)); - }else{ alignment = NULL; } + 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; - int minRGroup = -1; - int minRPos = 0; + vector< vector > minRGroup; + vector minRPos; - for(map::iterator it=ibarcodes.begin();it!=ibarcodes.end();it++){ - string oligo = it->second.reverse; - - if(rawRSequence.length() < maxLength){ //let's just assume that the barcodes are the same length + 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((rawRSequence.length()-(oligo.length()+bdiffs)),oligo.length()+bdiffs)); + alignment->alignPrimer(oligo, rawRSequence.substr(0,oligo.length()+bdiffs)); oligo = alignment->getSeqAAln(); string temp = alignment->getSeqBAln(); int alnLength = oligo.length(); - for(int i=0;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 = it->first; - minRPos = 0; + minRGroup.clear(); + minRGroup.push_back(it->second); + int tempminRPos = 0; + minRPos.clear(); for(int i=0;isecond); } } - - if(minDiff > bdiffs) { success = minDiff; } //no good matches - else if(minCount > 1) { success = bdiffs + 100; } //can't tell the difference between multiple barcodes - else{ + + /*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) { trashCode += "d"; 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 (minFGroup == minRGroup) { - group = minFGroup; - forwardSeq.setUnaligned(rawFSequence.substr(minFPos)); - reverseSeq.setUnaligned(rawRSequence.substr(0,(rawRSequence.length()-minRPos))); - forwardQual.trimQScores(minFPos, -1); - reverseQual.trimQScores(-1, rawRSequence.length()-minRPos); + if (foundMatch) { + forwardSeq.setUnaligned(rawFSequence.substr(fStart)); + reverseSeq.setUnaligned(rawRSequence.substr(rStart)); success = minDiff; - }else { success = bdiffs + 100; } + }else { if (trashCode == "") { trashCode = "bd"; } success = bdiffs + 100; } } - } + //} if (alignment != NULL) { delete alignment; } } + //catchall for case I didn't think of + if ((trashCode == "") && (success > bdiffs)) { trashCode = "bd"; } + return success; } @@ -366,255 +655,1222 @@ int TrimOligos::stripBarcode(Sequence& forwardSeq, Sequence& reverseSeq, Quality } //*******************************************************************/ -int TrimOligos::stripBarcode(Sequence& seq, int& group){ +int TrimOligos::stripBarcode(Sequence& forwardSeq, Sequence& reverseSeq, QualityScores& forwardQual, QualityScores& reverseQual, int& group){ try { - - string rawSequence = seq.getUnaligned(); + trashCode = ""; + + //look for forward barcode + string rawFSequence = forwardSeq.getUnaligned(); + string rawRSequence = reverseSeq.getUnaligned(); int success = bdiffs + 1; //guilty until proven innocent - //can you find the barcode - for(map::iterator it=barcodes.begin();it!=barcodes.end();it++){ - string oligo = it->first; - if(rawSequence.length() < oligo.length()){ //let's just assume that the barcodes are the same length + //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 + trashCode = "b"; + 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 + trashCode += "d"; break; } - if(compareDNASeq(oligo, rawSequence.substr(0,oligo.length()))){ - group = it->second; - seq.setUnaligned(rawSequence.substr(oligo.length())); - - success = 0; - break; + if(compareDNASeq(foligo, rawFSequence.substr(0,foligo.length()))) { + if (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; trashCode = ""; + return success; + } + }else { + trashCode = "b"; + if (!compareDNASeq(roligo, rawRSequence.substr((rawRSequence.length()-roligo.length()),roligo.length()))) { trashCode += "d"; + } } } //if you found the barcode or if you don't want to allow for diffs - if ((bdiffs == 0) || (success == 0)) { return success; } - + if (bdiffs == 0) { return success; } else { //try aligning and see if you can find it - - int maxLength = 0; - + trashCode = ""; 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; } + 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; - int minGroup = -1; - int minPos = 0; - - for(map::iterator it=barcodes.begin();it!=barcodes.end();it++){ + 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; - // int length = oligo.length(); - if(rawSequence.length() < maxLength){ //let's just assume that the barcodes are the same length + 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, rawSequence.substr(0,oligo.length()+bdiffs)); + alignment->alignPrimer(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; } - } + 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); + 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; - minGroup = it->second; - minPos = 0; + 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) { trashCode = "b"; minFGroup.clear(); 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->alignPrimer(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) { trashCode = "d"; 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 { if (trashCode == "") { trashCode = "bd"; } success = bdiffs + 100; } + } + //} - if(minDiff > bdiffs) { success = minDiff; } //no good matches - else if(minCount > 1) { success = bdiffs + 100; } //can't tell the difference between multiple barcodes - else{ //use the best match - group = minGroup; - seq.setUnaligned(rawSequence.substr(minPos)); - success = minDiff; + if (alignment != NULL) { delete alignment; } + } + + //catchall for case I didn't think of + if ((trashCode == "") && (success > bdiffs)) { trashCode = "bd"; } + + return success; + + } + catch(exception& e) { + m->errorOut(e, "TrimOligos", "stripIBarcode"); + exit(1); + } + +} +//*******************************************************************/ +int TrimOligos::stripPairedBarcode(Sequence& seq, QualityScores& qual, int& group){ + try { + trashCode = ""; + + //look for forward barcode + string rawSeq = seq.getUnaligned(); + int success = bdiffs + 1; //guilty until proven innocent + //cout << seq.getName() << endl; + //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; + //cout << it->first << '\t' << foligo << '\t' << roligo << endl; + //cout << it->first << '\t' << rawSeq.substr(0,foligo.length()) << '\t' << rawSeq.substr(rawSeq.length()-roligo.length(),roligo.length()) << endl; + if(rawSeq.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 + trashCode = "b"; + break; + } + if(rawSeq.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 + trashCode += "d"; + break; + } + + if(compareDNASeq(foligo, rawSeq.substr(0,foligo.length()))) { //find forward + if (compareDNASeq(roligo, rawSeq.substr(rawSeq.length()-roligo.length(),roligo.length()))) { //find reverse + group = it->first; + string trimmedSeq = rawSeq.substr(foligo.length()); //trim forward barcode + seq.setUnaligned(trimmedSeq.substr(0,(trimmedSeq.length()-roligo.length()))); //trim reverse barcode + if(qual.getName() != ""){ + qual.trimQScores(-1, rawSeq.length()-roligo.length()); + qual.trimQScores(foligo.length(), -1); + } + success = 0; + trashCode = ""; + return success; + } + }else { + trashCode = "b"; + if (!compareDNASeq(roligo, rawSeq.substr(rawSeq.length()-roligo.length(),roligo.length()))) { trashCode += "d"; } + } + } + //cout << "success=" << success << endl; + //if you found the barcode or if you don't want to allow for diffs + if (bdiffs == 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(rawSeq.length() < maxFBarcodeLength){ //let's just assume that the barcodes are the same length + success = bdiffs + 10; + break; + } + //cout << "before = " << oligo << '\t' << rawSeq.substr(0,oligo.length()+bdiffs) << endl; + //use needleman to align first barcode.length()+numdiffs of sequence to each barcode + alignment->alignPrimer(oligo, rawSeq.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) { trashCode = "b"; minFGroup.clear(); 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; + + string rawRSequence = reverseOligo(seq.getUnaligned()); + //cout << irbarcodes.size() << '\t' << maxRBarcodeLength << endl; + for(map >::iterator it=irbarcodes.begin();it!=irbarcodes.end();it++){ + string oligo = reverseOligo(it->first); + //cout << "r before = " << reverseOligo(oligo) << '\t' << reverseOligo(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->alignPrimer(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 << "r after = " << reverseOligo(oligo) << '\t' << reverseOligo(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) { trashCode += "d"; 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 = rawSeq.length() - minRPos[0]; + } + + //we have an acceptable match for the forward and reverse, but do they match? + if (foundMatch) { + string trimmedSeq = rawSeq.substr(0, rStart); //trim reverse barcode + seq.setUnaligned(trimmedSeq.substr(fStart)); //trim forward barcode + if(qual.getName() != ""){ + qual.trimQScores(-1, rStart); + qual.trimQScores(fStart, -1); + } + success = minDiff; + //cout << "barcode = " << ipbarcodes[group].forward << '\t' << ipbarcodes[group].reverse << endl; + }else { if (trashCode == "") { trashCode = "bd"; } success = bdiffs + 100; } + } + //} + + if (alignment != NULL) { delete alignment; } + } + + //catchall for case I didn't think of + if ((trashCode == "") && (success > bdiffs)) { trashCode = "bd"; } + + return success; + + } + catch(exception& e) { + m->errorOut(e, "TrimOligos", "stripPairedBarcode"); + exit(1); + } + +} +//*******************************************************************/ +int TrimOligos::stripPairedPrimers(Sequence& seq, QualityScores& qual, int& group, bool keepForward){ + try { + trashCode = ""; + //look for forward + string rawSeq = seq.getUnaligned(); + int success = pdiffs + 1; //guilty until proven innocent + //cout << seq.getName() << endl; + //can you find the forward + for(map::iterator it=ipprimers.begin();it!=ipprimers.end();it++){ + string foligo = it->second.forward; + string roligo = it->second.reverse; + + //cout << it->first << '\t' << foligo << '\t' << roligo << endl; + //cout << it->first << '\t' << rawSeq.substr(0,foligo.length()) << '\t' << rawSeq.substr(rawSeq.length()-roligo.length(),roligo.length()) << endl; + if(rawSeq.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 + trashCode = "p"; + break; + } + if(rawSeq.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 + trashCode += "q"; + break; + } + + if(compareDNASeq(foligo, rawSeq.substr(0,foligo.length()))) { //find forward + if (compareDNASeq(roligo, rawSeq.substr(rawSeq.length()-roligo.length(),roligo.length()))) { //find reverse + group = it->first; + string trimmedSeq = rawSeq.substr(foligo.length()); //trim forward barcode + seq.setUnaligned(trimmedSeq.substr(0,(trimmedSeq.length()-roligo.length()))); //trim reverse barcode + if(qual.getName() != ""){ + qual.trimQScores(-1, rawSeq.length()-roligo.length()); + qual.trimQScores(foligo.length(), -1); + } + success = 0; + trashCode = ""; + return success; + } + }else { + trashCode = "b"; + if (!compareDNASeq(roligo, rawSeq.substr(rawSeq.length()-roligo.length(),roligo.length()))) { trashCode += "d"; } + } + + } + //cout << "success=" << success << endl; + //if you found the barcode or if you don't want to allow for diffs + if (pdiffs == 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(rawSeq.length() < maxFPrimerLength){ //let's just assume that the barcodes are the same length + success = pdiffs + 10; + break; + } + //cout << "before = " << oligo << '\t' << rawSeq.substr(0,oligo.length()+bdiffs) << endl; + //use needleman to align first barcode.length()+numdiffs of sequence to each barcode + alignment->alignPrimer(oligo, rawSeq.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) { trashCode = "p"; minFGroup.clear(); success = minDiff; } //no good matches + //else{ + //check for reverse match + if (alignment != NULL) { delete alignment; } + + if (irprimers.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; + + string rawRSequence = reverseOligo(seq.getUnaligned()); + + for(map >::iterator it=irprimers.begin();it!=irprimers.end();it++){ + string oligo = reverseOligo(it->first); + //cout << "r before = " << reverseOligo(oligo) << '\t' << reverseOligo(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->alignPrimer(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 << "r after = " << reverseOligo(oligo) << '\t' << reverseOligo(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) { trashCode += "q"; 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 = rawSeq.length() - minRPos[0]; + } + + //we have an acceptable match for the forward and reverse, but do they match? + if (foundMatch) { + if (!keepForward) { + string trimmedSeq = rawSeq.substr(0, rStart); //trim reverse barcode + seq.setUnaligned(trimmedSeq.substr(fStart)); //trim forward barcode + if(qual.getName() != ""){ + qual.trimQScores(-1, rStart); + qual.trimQScores(fStart, -1); + } + } + success = minDiff; + //cout << "primer = " << ipprimers[group].forward << '\t' << ipprimers[group].reverse << endl; + }else { if (trashCode == "") { trashCode = "pq"; } success = pdiffs + 100; } + } + //} + + if (alignment != NULL) { delete alignment; } + } + + //catchall for case I didn't think of + if ((trashCode == "") && (success > pdiffs)) { trashCode = "pq"; } + + return success; + + } + catch(exception& e) { + m->errorOut(e, "TrimOligos", "stripPairedPrimers"); + 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->alignPrimer(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->alignPrimer(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", "stripBarcode"); + m->errorOut(e, "TrimOligos", "stripIForward"); exit(1); } } -/******************************************************************* -int TrimOligos::stripRBarcode(Sequence& seq, QualityScores& qual, int& group){ +//*******************************************************************/ +int TrimOligos::stripForward(Sequence& forwardSeq, Sequence& reverseSeq, int& group){ try { + //look for forward barcode + string rawFSequence = forwardSeq.getUnaligned(); + string rawRSequence = reverseSeq.getUnaligned(); + int success = pdiffs + 1; //guilty until proven innocent - string rawSequence = seq.getUnaligned(); - int success = bdiffs + 1; //guilty until proven innocent - - //can you find the barcode - for(map::iterator it=rbarcodes.begin();it!=rbarcodes.end();it++){ - string oligo = it->first; - if(rawSequence.length() < oligo.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 + //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(compareDNASeq(oligo, rawSequence.substr((rawSequence.length()-oligo.length()),oligo.length()))){ - group = it->second; - seq.setUnaligned(rawSequence.substr(0,(rawSequence.length()-oligo.length()))); - - if(qual.getName() != ""){ - qual.trimQScores(-1, rawSequence.length()-oligo.length()); - } - + 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()))); 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; } - + if ((pdiffs == 0) || (success == 0)) { return success; } else { //try aligning and see if you can find it - - int maxLength = 0; - Alignment* alignment; - if (rbarcodes.size() > 0) { - map::iterator it; - - for(it=rbarcodes.begin();it!=rbarcodes.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 (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; - int minGroup = -1; - int minPos = 0; - - for(map::iterator it=rbarcodes.begin();it!=rbarcodes.end();it++){ + 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; - // int length = oligo.length(); - if(rawSequence.length() < maxLength){ //let's just assume that the barcodes are the same length - success = bdiffs + 10; + 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, rawSequence.substr((rawSequence.length()-(oligo.length()+bdiffs)),oligo.length()+bdiffs)); + alignment->alignPrimer(oligo, rawFSequence.substr(0,oligo.length()+pdiffs)); oligo = alignment->getSeqAAln(); string temp = alignment->getSeqBAln(); - + int alnLength = oligo.length(); - for(int i=0;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; - minGroup = it->second; - minPos = 0; - for(int i=alnLength-1;i>=0;i--){ + 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 if(minCount > 1) { success = bdiffs + 100; } //can't tell the difference between multiple barcodes - else{ //use the best match - group = minGroup; - seq.setUnaligned(rawSequence.substr(0, (rawSequence.length()-minPos))); + + //cout << minDiff << '\t' << minCount << '\t' << endl; + if(minDiff > pdiffs) { success = minDiff; } //no good matches + else{ + //check for reverse match + if (alignment != NULL) { delete alignment; } - if(qual.getName() != ""){ - qual.trimQScores(-1, (rawSequence.length()-minPos)); - } - success = minDiff; + 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->alignPrimer(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)); + success = minDiff; + }else { success = pdiffs + 100; } + } } if (alignment != NULL) { delete alignment; } - } return success; } catch(exception& e) { - m->errorOut(e, "TrimOligos", "stripRBarcode"); + m->errorOut(e, "TrimOligos", "stripIForward"); exit(1); } } -/******************************************************************* -int TrimOligos::stripRBarcode(Sequence& seq, int& group){ +//*******************************************************************/ +int TrimOligos::stripBarcode(Sequence& seq, int& group){ try { string rawSequence = seq.getUnaligned(); int success = bdiffs + 1; //guilty until proven innocent //can you find the barcode - for(map::iterator it=rbarcodes.begin();it!=rbarcodes.end();it++){ + for(map::iterator it=barcodes.begin();it!=barcodes.end();it++){ string oligo = it->first; if(rawSequence.length() < oligo.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(oligo, rawSequence.substr((rawSequence.length()-oligo.length()),oligo.length()))){ + + if(compareDNASeq(oligo, rawSequence.substr(0,oligo.length()))){ group = it->second; - seq.setUnaligned(rawSequence.substr(0,(rawSequence.length()-oligo.length()))); + seq.setUnaligned(rawSequence.substr(oligo.length())); success = 0; break; @@ -625,21 +1881,9 @@ int TrimOligos::stripRBarcode(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 (rbarcodes.size() > 0) { - map::iterator it; - - for(it=rbarcodes.begin();it!=rbarcodes.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; @@ -647,28 +1891,28 @@ int TrimOligos::stripRBarcode(Sequence& seq, int& group){ int minGroup = -1; int minPos = 0; - for(map::iterator it=rbarcodes.begin();it!=rbarcodes.end();it++){ + for(map::iterator it=barcodes.begin();it!=barcodes.end();it++){ 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; } //use needleman to align first barcode.length()+numdiffs of sequence to each barcode - alignment->align(oligo, rawSequence.substr((rawSequence.length()-(oligo.length()+bdiffs)),oligo.length()+bdiffs)); + alignment->alignPrimer(oligo, rawSequence.substr(0,oligo.length()+bdiffs)); oligo = alignment->getSeqAAln(); string temp = alignment->getSeqBAln(); - + int alnLength = oligo.length(); - for(int i=0;i=0;i--){ + if(oligo[i] != '-'){ alnLength = i+1; break; } } - oligo = oligo.substr(alnLength); - temp = temp.substr(alnLength); - + oligo = oligo.substr(0,alnLength); + temp = temp.substr(0,alnLength); + int numDiff = countDiffs(oligo, temp); if(numDiff < minDiff){ @@ -676,7 +1920,7 @@ int TrimOligos::stripRBarcode(Sequence& seq, int& group){ minCount = 1; minGroup = it->second; minPos = 0; - for(int i=alnLength-1;i>=0;i--){ + for(int i=0;i 1) { success = bdiffs + 100; } //can't tell the difference between multiple barcodes else{ //use the best match group = minGroup; - seq.setUnaligned(rawSequence.substr(0, (rawSequence.length()-minPos))); - + seq.setUnaligned(rawSequence.substr(minPos)); success = minDiff; } @@ -705,12 +1948,13 @@ int TrimOligos::stripRBarcode(Sequence& seq, int& group){ } catch(exception& e) { - m->errorOut(e, "TrimOligos", "stripRBarcode"); + m->errorOut(e, "TrimOligos", "stripBarcode"); exit(1); } } -//********************************************************************/ + +/********************************************************************/ int TrimOligos::stripForward(Sequence& seq, int& group){ try { @@ -737,21 +1981,9 @@ 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; @@ -763,13 +1995,13 @@ 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; } //use needleman to align first barcode.length()+numdiffs of sequence to each barcode - alignment->align(oligo, rawSequence.substr(0,oligo.length()+pdiffs)); + alignment->alignPrimer(oligo, rawSequence.substr(0,oligo.length()+pdiffs)); oligo = alignment->getSeqAAln(); string temp = alignment->getSeqBAln(); @@ -823,6 +2055,9 @@ int TrimOligos::stripForward(Sequence& seq, int& group){ //*******************************************************************/ int TrimOligos::stripForward(Sequence& seq, QualityScores& qual, int& group, bool keepForward){ try { + + if (paired) { int success = stripPairedPrimers(seq, qual, group, keepForward); return success; } + string rawSequence = seq.getUnaligned(); int success = pdiffs + 1; //guilty until proven innocent @@ -849,21 +2084,9 @@ 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; @@ -875,13 +2098,13 @@ 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; } //use needleman to align first barcode.length()+numdiffs of sequence to each barcode - alignment->align(oligo, rawSequence.substr(0,oligo.length()+pdiffs)); + alignment->alignPrimer(oligo, rawSequence.substr(0,oligo.length()+pdiffs)); oligo = alignment->getSeqAAln(); string temp = alignment->getSeqBAln(); @@ -1024,19 +2247,9 @@ bool TrimOligos::stripLinker(Sequence& seq, QualityScores& qual){ if ((ldiffs == 0) || (success == 0)) { return success; } else { //try aligning and see if you can find it - - int maxLength = 0; - Alignment* alignment; - if (linker.size() > 0) { - for(int i = 0; i < linker.size(); i++){ - if(linker[i].length() > maxLength){ - maxLength = linker[i].length(); - } - } - alignment = new NeedlemanOverlap(-1.0, 1.0, -1.0, (maxLength+ldiffs+1)); - - }else{ alignment = NULL; } + if (linker.size() > 0) { alignment = new NeedlemanOverlap(-1.0, 1.0, -1.0, (maxLinkerLength+ldiffs+1)); } + else{ alignment = NULL; } //can you find the barcode int minDiff = 1e6; @@ -1047,13 +2260,13 @@ bool TrimOligos::stripLinker(Sequence& seq, QualityScores& qual){ string oligo = linker[i]; // int length = oligo.length(); - if(rawSequence.length() < maxLength){ //let's just assume that the barcodes are the same 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)); + alignment->alignPrimer(oligo, rawSequence.substr(0,oligo.length()+ldiffs)); oligo = alignment->getSeqAAln(); string temp = alignment->getSeqBAln(); @@ -1133,19 +2346,9 @@ bool TrimOligos::stripLinker(Sequence& seq){ if ((ldiffs == 0) || (success == 0)) { return success; } else { //try aligning and see if you can find it - - int maxLength = 0; - Alignment* alignment; - if (linker.size() > 0) { - for(int i = 0; i < linker.size(); i++){ - if(linker[i].length() > maxLength){ - maxLength = linker[i].length(); - } - } - alignment = new NeedlemanOverlap(-1.0, 1.0, -1.0, (maxLength+ldiffs+1)); - - }else{ alignment = NULL; } + if (linker.size() > 0) {alignment = new NeedlemanOverlap(-1.0, 1.0, -1.0, (maxLinkerLength+ldiffs+1)); } + else{ alignment = NULL; } //can you find the barcode int minDiff = 1e6; @@ -1156,13 +2359,13 @@ bool TrimOligos::stripLinker(Sequence& seq){ string oligo = linker[i]; // int length = oligo.length(); - if(rawSequence.length() < maxLength){ //let's just assume that the barcodes are the same 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)); + alignment->alignPrimer(oligo, rawSequence.substr(0,oligo.length()+ldiffs)); oligo = alignment->getSeqAAln(); string temp = alignment->getSeqBAln(); @@ -1240,19 +2443,9 @@ bool TrimOligos::stripSpacer(Sequence& seq, QualityScores& qual){ if ((sdiffs == 0) || (success == 0)) { return success; } else { //try aligning and see if you can find it - - int maxLength = 0; - Alignment* alignment; - if (spacer.size() > 0) { - for(int i = 0; i < spacer.size(); i++){ - if(spacer[i].length() > maxLength){ - maxLength = spacer[i].length(); - } - } - alignment = new NeedlemanOverlap(-1.0, 1.0, -1.0, (maxLength+sdiffs+1)); - - }else{ alignment = NULL; } + if (spacer.size() > 0) { alignment = new NeedlemanOverlap(-1.0, 1.0, -1.0, (maxSpacerLength+sdiffs+1)); } + else{ alignment = NULL; } //can you find the barcode int minDiff = 1e6; @@ -1263,13 +2456,13 @@ bool TrimOligos::stripSpacer(Sequence& seq, QualityScores& qual){ string oligo = spacer[i]; // int length = oligo.length(); - if(rawSequence.length() < maxLength){ //let's just assume that the barcodes are the same 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)); + alignment->alignPrimer(oligo, rawSequence.substr(0,oligo.length()+sdiffs)); oligo = alignment->getSeqAAln(); string temp = alignment->getSeqBAln(); @@ -1349,19 +2542,9 @@ bool TrimOligos::stripSpacer(Sequence& seq){ if ((sdiffs == 0) || (success == 0)) { return success; } else { //try aligning and see if you can find it - - int maxLength = 0; - Alignment* alignment; - if (spacer.size() > 0) { - for(int i = 0; i < spacer.size(); i++){ - if(spacer[i].length() > maxLength){ - maxLength = spacer[i].length(); - } - } - alignment = new NeedlemanOverlap(-1.0, 1.0, -1.0, (maxLength+sdiffs+1)); - - }else{ alignment = NULL; } + if (spacer.size() > 0) { alignment = new NeedlemanOverlap(-1.0, 1.0, -1.0, (maxSpacerLength+sdiffs+1)); } + else{ alignment = NULL; } //can you find the barcode int minDiff = 1e6; @@ -1372,13 +2555,13 @@ bool TrimOligos::stripSpacer(Sequence& seq){ string oligo = spacer[i]; // int length = oligo.length(); - if(rawSequence.length() < maxLength){ //let's just assume that the barcodes are the same 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)); + alignment->alignPrimer(oligo, rawSequence.substr(0,oligo.length()+sdiffs)); oligo = alignment->getSeqAAln(); string temp = alignment->getSeqBAln(); @@ -1498,6 +2681,48 @@ int TrimOligos::countDiffs(string oligo, string seq){ exit(1); } } +//********************************************************************/ +string TrimOligos::reverseOligo(string oligo){ + try { + string reverse = ""; + + for(int i=oligo.length()-1;i>=0;i--){ + + if(oligo[i] == 'A') { reverse += 'T'; } + else if(oligo[i] == 'T'){ reverse += 'A'; } + else if(oligo[i] == 'U'){ reverse += 'A'; } + + else if(oligo[i] == 'G'){ reverse += 'C'; } + else if(oligo[i] == 'C'){ reverse += 'G'; } + + else if(oligo[i] == 'R'){ reverse += 'Y'; } + else if(oligo[i] == 'Y'){ reverse += 'R'; } + + else if(oligo[i] == 'M'){ reverse += 'K'; } + else if(oligo[i] == 'K'){ reverse += 'M'; } + + else if(oligo[i] == 'W'){ reverse += 'W'; } + else if(oligo[i] == 'S'){ reverse += 'S'; } + + else if(oligo[i] == 'B'){ reverse += 'V'; } + else if(oligo[i] == 'V'){ reverse += 'B'; } + + else if(oligo[i] == 'D'){ reverse += 'H'; } + else if(oligo[i] == 'H'){ reverse += 'D'; } + + else if(oligo[i] == '-'){ reverse += '-'; } + else { reverse += 'N'; } + } + + + return reverse; + } + catch(exception& e) { + m->errorOut(e, "TrimOligos", "reverseOligo"); + exit(1); + } +} + /********************************************************************/