makecontigscommand.h

   1 #ifndef Mothur_makecontigscommand_h
   2 #define Mothur_makecontigscommand_h
   3
   4 //
   5 //  makecontigscommand.h
   6 //  Mothur
   7 //
   8 //  Created by Sarah Westcott on 5/15/12.
   9 //  Copyright (c) 2012 Schloss Lab. All rights reserved.
  10 //
  11
  12 #include "command.hpp"
  13 #include "sequence.hpp"
  14 #include "qualityscores.h"
  15 #include "alignment.hpp"
  16 #include "gotohoverlap.hpp"
  17 #include "needlemanoverlap.hpp"
  18 #include "blastalign.hpp"
  19 #include "noalign.hpp"
  20 #include "trimoligos.h"
  21
  22 struct fastqRead {
  23         vector<int> scores;
  24         string name;
  25         string sequence;
  26
  27         fastqRead() { name = ""; sequence = ""; scores.clear(); };
  28         fastqRead(string n, string s, vector<int> sc) : name(n), sequence(s), scores(sc) {};
  29         ~fastqRead() {};
  30 };
  31
  32 struct pairFastqRead {
  33         fastqRead forward;
  34     fastqRead reverse;
  35
  36         pairFastqRead() {};
  37         pairFastqRead(fastqRead f, fastqRead r) : forward(f), reverse(r){};
  38         ~pairFastqRead() {};
  39 };
  40 /**************************************************************************************************/
  41
  42 class MakeContigsCommand : public Command {
  43 public:
  44     MakeContigsCommand(string);
  45     MakeContigsCommand();
  46     ~MakeContigsCommand(){}
  47
  48     vector<string> setParameters();
  49     string getCommandName()                     { return "make.contigs";                        }
  50     string getCommandCategory()         { return "Sequence Processing";         }
  51     //commmand category choices: Sequence Processing, OTU-Based Approaches, Hypothesis Testing, Phylotype Analysis, General, Clustering and Hidden
  52
  53         string getHelpString();
  54     string getOutputPattern(string);
  55     string getCitation() { return "http://www.mothur.org/wiki/Make.contigs"; }
  56     string getDescription()             { return "description"; }
  57
  58     int execute();
  59     void help() { m->mothurOut(getHelpString()); }
  60
  61 private:
  62     bool abort, allFiles, createGroup;
  63     string outputDir, ffastqfile, rfastqfile, align, oligosfile;
  64         float match, misMatch, gapOpen, gapExtend;
  65         int processors, longestBase, threshold, tdiffs, bdiffs, pdiffs, ldiffs, sdiffs;
  66     vector<string> outputNames;
  67
  68     map<int, oligosPair> barcodes;
  69         map<int, oligosPair> primers;
  70     vector<string>  linker;
  71     vector<string>  spacer;
  72         vector<string> primerNameVector;
  73         vector<string> barcodeNameVector;
  74
  75         map<string, int> groupCounts;
  76     map<string, string> groupMap;
  77     //map<string, int> combos;
  78         //map<string, int> groupToIndex;
  79     //vector<string> groupVector;
  80
  81     fastqRead readFastq(ifstream&, bool&);
  82     vector< vector<string> > readFastqFiles(int&);
  83     bool checkReads(fastqRead&, fastqRead&);
  84     int createProcesses(vector< vector<string> >, string, string, string, string, string, vector<vector<string> >, vector<vector<string> >);
  85     int driver(vector<string>, string, string, string, string, string, vector<vector<string> >, vector<vector<string> >);
  86     bool getOligos(vector<vector<string> >&, vector<vector<string> >&);
  87     string reverseOligo(string);
  88     vector<pairFastqRead> getReads(bool ignoref, bool ignorer, fastqRead forward, fastqRead reverse, map<string, fastqRead>& uniques);
  89 };
  90
  91 /**************************************************************************************************/
  92
  93 /**************************************************************************************************/
  94 //custom data structure for threads to use.
  95 // This is passed by void pointer so it can be any data type
  96 // that can be passed using a single void pointer (LPVOID).
  97 struct contigsData {
  98         string outputFasta;
  99         string outputQual;
 100     string outputScrapFasta;
 101         string outputScrapQual;
 102         string outputMisMatches;
 103         string align;
 104     vector<string> files;
 105     vector<vector<string> > fastaFileNames;
 106     vector<vector<string> > qualFileNames;
 107         MothurOut* m;
 108         float match, misMatch, gapOpen, gapExtend;
 109         int count, threshold, threadID, pdiffs, bdiffs, tdiffs;
 110     bool allFiles, createGroup;
 111     map<string, int> groupCounts;
 112     map<string, string> groupMap;
 113     vector<string> primerNameVector;
 114         vector<string> barcodeNameVector;
 115     map<int, oligosPair> barcodes;
 116         map<int, oligosPair> primers;
 117
 118         contigsData(){}
 119         contigsData(vector<string> f, string of, string oq, string osf, string osq, string om, string al, MothurOut* mout, float ma, float misMa, float gapO, float gapE, int thr, map<int, oligosPair> br, map<int, oligosPair> pr, vector<vector<string> > ffn, vector<vector<string> > qfn, vector<string>bnv, vector<string> pnv, int pdf, int bdf, int tdf, bool cg, bool all, int tid) {
 120         files = f;
 121                 outputFasta = of;
 122         outputQual = oq;
 123         outputMisMatches = om;
 124         m = mout;
 125                 match = ma;
 126                 misMatch = misMa;
 127                 gapOpen = gapO;
 128                 gapExtend = gapE;
 129         threshold = thr;
 130                 align = al;
 131                 count = 0;
 132         outputScrapFasta = osf;
 133         outputScrapQual = osq;
 134         fastaFileNames = ffn;
 135         qualFileNames = qfn;
 136         barcodes = br;
 137         primers = pr;
 138         barcodeNameVector = bnv;
 139         primerNameVector = pnv;
 140         pdiffs = pdf;
 141         bdiffs = bdf;
 142         tdiffs = tdf;
 143         allFiles = all;
 144         createGroup = cg;
 145                 threadID = tid;
 146         }
 147 };
 148
 149 /**************************************************************************************************/
 150 #if defined (__APPLE__) || (__MACH__) || (linux) || (__linux) || (__linux__) || (__unix__) || (__unix)
 151 #else
 152 static DWORD WINAPI MyContigsThreadFunction(LPVOID lpParam){
 153         contigsData* pDataArray;
 154         pDataArray = (contigsData*)lpParam;
 155
 156         try {
 157         int longestBase = 1000;
 158         Alignment* alignment;
 159         if(pDataArray->align == "gotoh")                        {       alignment = new GotohOverlap(pDataArray->gapOpen, pDataArray->gapExtend, pDataArray->match, pDataArray->misMatch, longestBase);                 }
 160                 else if(pDataArray->align == "needleman")       {       alignment = new NeedlemanOverlap(pDataArray->gapOpen, pDataArray->match, pDataArray->misMatch, longestBase);                            }
 161
 162         int num = 0;
 163         string thisffastafile = pDataArray->files[0];
 164         string thisfqualfile = pDataArray->files[1];
 165         string thisrfastafile = pDataArray->files[2];
 166         string thisrqualfile = pDataArray->files[3];
 167
 168         if (pDataArray->m->debug) {  pDataArray->m->mothurOut("[DEBUG]: ffasta = " + thisffastafile + ".\n[DEBUG]: fqual = " + thisfqualfile + ".\n[DEBUG]: rfasta = " + thisrfastafile + ".\n[DEBUG]: rqual = " + thisrqualfile + ".\n"); }
 169
 170                 if(pDataArray->allFiles){
 171                         for (int i = 0; i < pDataArray->fastaFileNames.size(); i++) { //clears old file
 172                                 for (int j = 0; j < pDataArray->fastaFileNames[i].size(); j++) { //clears old file
 173                                         if (pDataArray->fastaFileNames[i][j] != "") {
 174                                                 ofstream temp;
 175                                                 pDataArray->m->openOutputFile(pDataArray->fastaFileNames[i][j], temp);                  temp.close();
 176                         pDataArray->m->openOutputFile(pDataArray->qualFileNames[i][j], temp);                   temp.close();
 177                                         }
 178                                 }
 179                         }
 180                 }
 181
 182         ifstream inFFasta, inRFasta, inFQual, inRQual;
 183         pDataArray->m->openInputFile(thisffastafile, inFFasta);
 184         pDataArray->m->openInputFile(thisfqualfile, inFQual);
 185         pDataArray->m->openInputFile(thisrfastafile, inRFasta);
 186         pDataArray->m->openInputFile(thisrqualfile, inRQual);
 187
 188         ofstream outFasta, outQual, outMisMatch, outScrapFasta, outScrapQual;
 189         pDataArray->m->openOutputFile(pDataArray->outputFasta, outFasta);
 190         pDataArray->m->openOutputFile(pDataArray->outputQual, outQual);
 191         pDataArray->m->openOutputFile(pDataArray->outputMisMatches, outMisMatch);
 192         pDataArray->m->openOutputFile(pDataArray->outputScrapFasta, outScrapFasta);
 193         pDataArray->m->openOutputFile(pDataArray->outputScrapQual, outScrapQual);
 194         outMisMatch << "Name\tLength\tMisMatches\n";
 195
 196         TrimOligos trimOligos(pDataArray->pdiffs, pDataArray->bdiffs, 0, 0, pDataArray->primers, pDataArray->barcodes);
 197
 198         while ((!inFQual.eof()) && (!inFFasta.eof()) && (!inRFasta.eof()) && (!inRQual.eof())) {
 199
 200             if (pDataArray->m->control_pressed) { break; }
 201
 202             int success = 1;
 203             string trashCode = "";
 204             int currentSeqsDiffs = 0;
 205
 206             //read seqs and quality info
 207             Sequence fSeq(inFFasta); pDataArray->m->gobble(inFFasta);
 208             Sequence rSeq(inRFasta); pDataArray->m->gobble(inRFasta);
 209             QualityScores fQual(inFQual); pDataArray->m->gobble(inFQual);
 210             QualityScores rQual(inRQual); pDataArray->m->gobble(inRQual);
 211
 212             int barcodeIndex = 0;
 213             int primerIndex = 0;
 214
 215             if(pDataArray->barcodes.size() != 0){
 216                 success = trimOligos.stripBarcode(fSeq, rSeq, fQual, rQual, barcodeIndex);
 217                 if(success > pDataArray->bdiffs)                {       trashCode += 'b';       }
 218                 else{ currentSeqsDiffs += success;  }
 219             }
 220
 221             if(pDataArray->primers.size() != 0){
 222                 success = trimOligos.stripForward(fSeq, rSeq, fQual, rQual, primerIndex);
 223                 if(success > pDataArray->pdiffs)                {       trashCode += 'f';       }
 224                 else{ currentSeqsDiffs += success;  }
 225             }
 226
 227             if (currentSeqsDiffs > pDataArray->tdiffs)  {       trashCode += 't';   }
 228
 229             //flip the reverse reads
 230             rSeq.reverseComplement();
 231             rQual.flipQScores();
 232
 233             //pairwise align
 234             alignment->align(fSeq.getUnaligned(), rSeq.getUnaligned());
 235             map<int, int> ABaseMap = alignment->getSeqAAlnBaseMap();
 236             map<int, int> BBaseMap = alignment->getSeqBAlnBaseMap();
 237             fSeq.setAligned(alignment->getSeqAAln());
 238             rSeq.setAligned(alignment->getSeqBAln());
 239             int length = fSeq.getAligned().length();
 240
 241             //traverse alignments merging into one contiguous seq
 242             string contig = "";
 243             vector<int> contigScores;
 244             int numMismatches = 0;
 245             string seq1 = fSeq.getAligned();
 246             string seq2 = rSeq.getAligned();
 247             vector<int> scores1 = fQual.getQualityScores();
 248             vector<int> scores2 = rQual.getQualityScores();
 249
 250             int overlapStart = fSeq.getStartPos();
 251             int seq2Start = rSeq.getStartPos();
 252             //bigger of the 2 starting positions is the location of the overlapping start
 253             if (overlapStart < seq2Start) { //seq2 starts later so take from 0 to seq2Start from seq1
 254                 overlapStart = seq2Start;
 255                 for (int i = 0; i < overlapStart; i++) {
 256                     contig += seq1[i];
 257                     contigScores.push_back(scores1[ABaseMap[i]]);
 258                 }
 259             }else { //seq1 starts later so take from 0 to overlapStart from seq2
 260                 for (int i = 0; i < overlapStart; i++) {
 261                     contig += seq2[i];
 262                     contigScores.push_back(scores2[BBaseMap[i]]);
 263                 }
 264             }
 265
 266             int seq1End = fSeq.getEndPos();
 267             int seq2End = rSeq.getEndPos();
 268             int overlapEnd = seq1End;
 269             if (seq2End < overlapEnd) { overlapEnd = seq2End; }  //smallest end position is where overlapping ends
 270
 271             for (int i = overlapStart; i < overlapEnd; i++) {
 272                 if (seq1[i] == seq2[i]) { //match, add base and choose highest score
 273                     contig += seq1[i];
 274                     contigScores.push_back(scores1[ABaseMap[i]]);
 275                     if (scores1[ABaseMap[i]] < scores2[BBaseMap[i]]) { contigScores[contigScores.size()-1] = scores2[BBaseMap[i]]; }
 276                 }else if (((seq1[i] == '.') || (seq1[i] == '-')) && ((seq2[i] != '-') && (seq2[i] != '.'))) { //seq1 is a gap and seq2 is a base, choose seq2, unless quality score for base is below threshold. In that case eliminate base
 277                     if (scores2[BBaseMap[i]] < pDataArray->threshold) { } //
 278                     else {
 279                         contig += seq2[i];
 280                         contigScores.push_back(scores2[BBaseMap[i]]);
 281                     }
 282                 }else if (((seq2[i] == '.') || (seq2[i] == '-')) && ((seq1[i] != '-') && (seq1[i] != '.'))) { //seq2 is a gap and seq1 is a base, choose seq1, unless quality score for base is below threshold. In that case eliminate base
 283                     if (scores1[ABaseMap[i]] < pDataArray->threshold) { } //
 284                     else {
 285                         contig += seq1[i];
 286                         contigScores.push_back(scores1[ABaseMap[i]]);
 287                     }
 288                 }else if (((seq1[i] != '-') && (seq1[i] != '.')) && ((seq2[i] != '-') && (seq2[i] != '.'))) { //both bases choose one with better quality
 289                     char c = seq1[i];
 290                     contigScores.push_back(scores1[ABaseMap[i]]);
 291                     if (scores1[ABaseMap[i]] < scores2[BBaseMap[i]]) { contigScores[contigScores.size()-1] = scores2[BBaseMap[i]]; c = seq2[i]; }
 292                     contig += c;
 293                     numMismatches++;
 294                 }else { //should never get here
 295                     pDataArray->m->mothurOut("[ERROR]: case I didn't think of seq1 = " + toString(seq1[i]) + " and seq2 = " + toString(seq2[i]) + "\n");
 296                 }
 297             }
 298
 299             if (seq1End < seq2End) { //seq1 ends before seq2 so take from overlap to length from seq2
 300                 for (int i = overlapEnd; i < length; i++) {
 301                     contig += seq2[i];
 302                     contigScores.push_back(scores2[BBaseMap[i]]);
 303                 }
 304             }else { //seq2 ends before seq1 so take from overlap to length from seq1
 305                 for (int i = overlapEnd; i < length; i++) {
 306                     contig += seq1[i];
 307                     contigScores.push_back(scores1[ABaseMap[i]]);
 308                 }
 309
 310             }
 311
 312             if(trashCode.length() == 0){
 313                 if (pDataArray->createGroup) {
 314                     if(pDataArray->barcodes.size() != 0){
 315                         string thisGroup = pDataArray->barcodeNameVector[barcodeIndex];
 316                         if (pDataArray->primers.size() != 0) {
 317                             if (pDataArray->primerNameVector[primerIndex] != "") {
 318                                 if(thisGroup != "") {
 319                                     thisGroup += "." + pDataArray->primerNameVector[primerIndex];
 320                                 }else {
 321                                     thisGroup = pDataArray->primerNameVector[primerIndex];
 322                                 }
 323                             }
 324                         }
 325
 326                         if (pDataArray->m->debug) { pDataArray->m->mothurOut(", group= " + thisGroup + "\n"); }
 327
 328                         pDataArray->groupMap[fSeq.getName()] = thisGroup;
 329
 330                         map<string, int>::iterator it = pDataArray->groupCounts.find(thisGroup);
 331                         if (it == pDataArray->groupCounts.end()) {      pDataArray->groupCounts[thisGroup] = 1; }
 332                         else { pDataArray->groupCounts[it->first] ++; }
 333
 334                     }
 335                 }
 336
 337                 if(pDataArray->allFiles){
 338                     ofstream output;
 339                     pDataArray->m->openOutputFileAppend(pDataArray->fastaFileNames[barcodeIndex][primerIndex], output);
 340                     output << ">" << fSeq.getName() << endl << contig << endl;
 341                     output.close();
 342
 343                     pDataArray->m->openOutputFileAppend(pDataArray->qualFileNames[barcodeIndex][primerIndex], output);
 344                     output << ">" << fSeq.getName() << endl;
 345                     for (int i = 0; i < contigScores.size(); i++) { output << contigScores[i] << ' '; }
 346                     output << endl;
 347                     output.close();
 348                 }
 349
 350                 //output
 351                 outFasta << ">" << fSeq.getName() << endl << contig << endl;
 352                 outQual << ">" << fSeq.getName() << endl;
 353                 for (int i = 0; i < contigScores.size(); i++) { outQual << contigScores[i] << ' '; }
 354                 outQual << endl;
 355                 outMisMatch << fSeq.getName() << '\t' << contig.length() << '\t' << numMismatches << endl;
 356             }else {
 357                 //output
 358                 outScrapFasta << ">" << fSeq.getName() << " | " << trashCode << endl << contig << endl;
 359                 outScrapQual << ">" << fSeq.getName() << " | " << trashCode << endl;
 360                 for (int i = 0; i < contigScores.size(); i++) { outScrapQual << contigScores[i] << ' '; }
 361                 outScrapQual << endl;
 362             }
 363             num++;
 364
 365                         //report progress
 366                         if((num) % 1000 == 0){  pDataArray->m->mothurOut(toString(num)); pDataArray->m->mothurOutEndLine();             }
 367                 }
 368
 369                 //report progress
 370                 if((num) % 1000 != 0){  pDataArray->m->mothurOut(toString(num)); pDataArray->m->mothurOutEndLine();             }
 371
 372         inFFasta.close();
 373         inFQual.close();
 374         inRFasta.close();
 375         inRQual.close();
 376         outFasta.close();
 377         outQual.close();
 378         outMisMatch.close();
 379         outScrapFasta.close();
 380         outScrapQual.close();
 381         delete alignment;
 382
 383         if (pDataArray->m->control_pressed) { pDataArray->m->mothurRemove(pDataArray->outputQual); pDataArray->m->mothurRemove(pDataArray->outputFasta);  pDataArray->m->mothurRemove(pDataArray->outputMisMatches); pDataArray->m->mothurRemove(pDataArray->outputScrapQual); pDataArray->m->mothurRemove(pDataArray->outputScrapFasta);}
 384
 385         return 0;
 386
 387         }
 388         catch(exception& e) {
 389                 pDataArray->m->errorOut(e, "AlignCommand", "MyContigsThreadFunction");
 390                 exit(1);
 391         }
 392 }
 393 #endif
 394
 395
 396 #endif