X-Git-Url: https://git.donarmstrong.com/?a=blobdiff_plain;f=trimoligos.cpp;h=a53f4e85388f0f7bbb6fda40c2074b4f555e29d3;hb=767c2ab76d70664af19b1afca8e1fd5233962698;hp=c35971a676f1c9af3de404bb5f0a4ca707a82f1f;hpb=5c5c0428f6d548c28a8b903ac80efed4f92d59db;p=mothur.git diff --git a/trimoligos.cpp b/trimoligos.cpp index c35971a..a53f4e8 100644 --- a/trimoligos.cpp +++ b/trimoligos.cpp @@ -144,6 +144,24 @@ TrimOligos::TrimOligos(int p, int b, map pr, map br, v barcodes = br; primers = pr; revPrimer = r; + + 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"); @@ -152,7 +170,135 @@ 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(); + + 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) { 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->align(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; return false; } //no good matches + else if(minCount > 1) { primerStart = 0; primerEnd = 0; return false; } //can't tell the difference between multiple primers + else{ return true; } + } + + 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 { + + 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; + } + + } + } + + 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 { @@ -259,7 +405,7 @@ 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 { //look for forward barcode string rawFSequence = forwardSeq.getUnaligned(); @@ -284,8 +430,6 @@ int TrimOligos::stripBarcode(Sequence& forwardSeq, Sequence& reverseSeq, Quality 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; } @@ -468,8 +612,6 @@ int TrimOligos::stripBarcode(Sequence& forwardSeq, Sequence& reverseSeq, Quality 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; } } @@ -488,24 +630,24 @@ int TrimOligos::stripBarcode(Sequence& forwardSeq, Sequence& reverseSeq, Quality } //*******************************************************************/ -int TrimOligos::stripForward(Sequence& forwardSeq, Sequence& reverseSeq, QualityScores& forwardQual, QualityScores& reverseQual, int& group){ +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 = pdiffs + 1; //guilty until proven innocent + int success = bdiffs + 1; //guilty until proven innocent //can you find the forward barcode - for(map::iterator it=ipprimers.begin();it!=ipprimers.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 = pdiffs + 10; //if the sequence is shorter than the barcode then bail out + 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 = pdiffs + 10; //if the sequence is shorter than the barcode then bail out + success = bdiffs + 10; //if the sequence is shorter than the barcode then bail out break; } @@ -521,10 +663,10 @@ int TrimOligos::stripForward(Sequence& forwardSeq, Sequence& reverseSeq, Quality } //if you found the barcode or if you don't want to allow for diffs - if ((pdiffs == 0) || (success == 0)) { return success; } + if ((bdiffs == 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)); } + if (ifbarcodes.size() > 0) { alignment = new NeedlemanOverlap(-1.0, 1.0, -1.0, (maxFBarcodeLength+bdiffs+1)); } else{ alignment = NULL; } //can you find the barcode @@ -549,16 +691,16 @@ int TrimOligos::stripForward(Sequence& forwardSeq, Sequence& reverseSeq, Quality 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++){ + for(map >::iterator it=ifbarcodes.begin();it!=ifbarcodes.end();it++){ string oligo = it->first; - if(rawFSequence.length() < maxFPrimerLength){ //let's just assume that the barcodes are the same length - success = pdiffs + 10; + 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()+pdiffs) << endl; + //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()+pdiffs)); + alignment->align(oligo, rawFSequence.substr(0,oligo.length()+bdiffs)); oligo = alignment->getSeqAAln(); string temp = alignment->getSeqBAln(); @@ -569,7 +711,7 @@ int TrimOligos::stripForward(Sequence& forwardSeq, Sequence& reverseSeq, Quality temp = temp.substr(0,alnLength); int numDiff = countDiffs(oligo, temp); - if (alnLength == 0) { numDiff = pdiffs + 100; } + if (alnLength == 0) { numDiff = bdiffs + 100; } //cout << "after = " << oligo << '\t' << temp << '\t' << numDiff << endl; if(numDiff < minDiff){ @@ -598,12 +740,12 @@ int TrimOligos::stripForward(Sequence& forwardSeq, Sequence& reverseSeq, Quality } //cout << minDiff << '\t' << minCount << '\t' << endl; - if(minDiff > pdiffs) { success = minDiff; } //no good matches + if(minDiff > 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, (maxRPrimerLength+pdiffs+1)); } + if (irbarcodes.size() > 0) { alignment = new NeedlemanOverlap(-1.0, 1.0, -1.0, (maxRBarcodeLength+bdiffs+1)); } else{ alignment = NULL; } //can you find the barcode @@ -612,16 +754,16 @@ int TrimOligos::stripForward(Sequence& forwardSeq, Sequence& reverseSeq, Quality vector< vector > minRGroup; vector minRPos; - for(map >::iterator it=irprimers.begin();it!=irprimers.end();it++){ + for(map >::iterator it=irbarcodes.begin();it!=irbarcodes.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; + //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()+pdiffs)); + alignment->align(oligo, rawRSequence.substr(0,oligo.length()+bdiffs)); oligo = alignment->getSeqAAln(); string temp = alignment->getSeqBAln(); @@ -630,7 +772,7 @@ int TrimOligos::stripForward(Sequence& forwardSeq, Sequence& reverseSeq, Quality oligo = oligo.substr(0,alnLength); temp = temp.substr(0,alnLength); int numDiff = countDiffs(oligo, temp); - if (alnLength == 0) { numDiff = pdiffs + 100; } + if (alnLength == 0) { numDiff = bdiffs + 100; } //cout << "after = " << oligo << '\t' << temp << '\t' << numDiff << endl; if(numDiff < minDiff){ @@ -672,7 +814,7 @@ int TrimOligos::stripForward(Sequence& forwardSeq, Sequence& reverseSeq, Quality cout << endl; } cout << endl;*/ - if(minDiff > pdiffs) { success = minDiff; } //no good matches + if(minDiff > bdiffs) { success = minDiff; } //no good matches else { bool foundMatch = false; vector matches; @@ -700,6 +842,460 @@ int TrimOligos::stripForward(Sequence& forwardSeq, Sequence& reverseSeq, Quality 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); + } + +} +//*******************************************************************/ +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 + + //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()))); + 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)); + success = minDiff; }else { success = pdiffs + 100; } } }