bayesian.cpp

   1 /*
   2  *  bayesian.cpp
   3  *  Mothur
   4  *
   5  *  Created by westcott on 11/3/09.
   6  *  Copyright 2009 Schloss Lab. All rights reserved.
   7  *
   8  */
   9
  10 #include "bayesian.h"
  11 #include "kmer.hpp"
  12 #include "phylosummary.h"
  13 #include "referencedb.h"
  14 /**************************************************************************************************/
  15 Bayesian::Bayesian(string tfile, string tempFile, string method, int ksize, int cutoff, int i, int tid, bool f) :
  16 Classify(), kmerSize(ksize), confidenceThreshold(cutoff), iters(i) {
  17         try {
  18                 ReferenceDB* rdb = ReferenceDB::getInstance();
  19
  20                 threadID = tid;
  21                 flip = f;
  22                 string baseName = tempFile;
  23
  24                 if (baseName == "saved") { baseName = rdb->getSavedReference(); }
  25
  26                 string baseTName = tfile;
  27                 if (baseTName == "saved") { baseTName = rdb->getSavedTaxonomy(); }
  28
  29                 /************calculate the probablity that each word will be in a specific taxonomy*************/
  30                 string tfileroot = baseTName.substr(0,baseTName.find_last_of(".")+1);
  31                 string tempfileroot = m->getRootName(m->getSimpleName(baseName));
  32                 string phyloTreeName = tfileroot + "tree.train";
  33                 string phyloTreeSumName = tfileroot + "tree.sum";
  34                 string probFileName = tfileroot + tempfileroot + char('0'+ kmerSize) + "mer.prob";
  35                 string probFileName2 = tfileroot + tempfileroot + char('0'+ kmerSize) + "mer.numNonZero";
  36
  37                 ofstream out;
  38                 ofstream out2;
  39
  40                 ifstream phyloTreeTest(phyloTreeName.c_str());
  41                 ifstream probFileTest2(probFileName2.c_str());
  42                 ifstream probFileTest(probFileName.c_str());
  43                 ifstream probFileTest3(phyloTreeSumName.c_str());
  44
  45                 int start = time(NULL);
  46
  47                 //if they are there make sure they were created after this release date
  48                 bool FilesGood = false;
  49                 if(probFileTest && probFileTest2 && phyloTreeTest && probFileTest3){
  50                         FilesGood = checkReleaseDate(probFileTest, probFileTest2, phyloTreeTest, probFileTest3);
  51                 }
  52
  53                 //if you want to save, but you dont need to calculate then just read
  54                 if (rdb->save && probFileTest && probFileTest2 && phyloTreeTest && probFileTest3 && FilesGood && (tempFile != "saved")) {
  55                         ifstream saveIn;
  56                         m->openInputFile(tempFile, saveIn);
  57
  58                         while (!saveIn.eof()) {
  59                                 Sequence temp(saveIn);
  60                                 m->gobble(saveIn);
  61
  62                                 rdb->referenceSeqs.push_back(temp);
  63                         }
  64                         saveIn.close();
  65                 }
  66
  67                 if(probFileTest && probFileTest2 && phyloTreeTest && probFileTest3 && FilesGood){
  68                         if (tempFile == "saved") { m->mothurOutEndLine();  m->mothurOut("Using sequences from " + rdb->getSavedReference() + " that are saved in memory.");     m->mothurOutEndLine(); }
  69
  70                         m->mothurOut("Reading template taxonomy...     "); cout.flush();
  71
  72                         phyloTree = new PhyloTree(phyloTreeTest, phyloTreeName);
  73
  74                         m->mothurOut("DONE."); m->mothurOutEndLine();
  75
  76                         genusNodes = phyloTree->getGenusNodes();
  77                         genusTotals = phyloTree->getGenusTotals();
  78
  79                         if (tfile == "saved") {
  80                                 m->mothurOutEndLine();  m->mothurOut("Using probabilties from " + rdb->getSavedTaxonomy() + " that are saved in memory...    ");        cout.flush();;
  81                                 wordGenusProb = rdb->wordGenusProb;
  82                                 WordPairDiffArr = rdb->WordPairDiffArr;
  83                         }else {
  84                                 m->mothurOut("Reading template probabilities...     "); cout.flush();
  85                                 readProbFile(probFileTest, probFileTest2, probFileName, probFileName2);
  86                         }
  87
  88                         //save probabilities
  89                         if (rdb->save) { rdb->wordGenusProb = wordGenusProb; rdb->WordPairDiffArr = WordPairDiffArr; }
  90                 }else{
  91
  92                         //create search database and names vector
  93                         generateDatabaseAndNames(tfile, tempFile, method, ksize, 0.0, 0.0, 0.0, 0.0);
  94
  95                         //prevents errors caused by creating shortcut files if you had an error in the sanity check.
  96                         if (m->control_pressed) {  m->mothurRemove(phyloTreeName);  m->mothurRemove(probFileName); m->mothurRemove(probFileName2); }
  97                         else{
  98                                 genusNodes = phyloTree->getGenusNodes();
  99                                 genusTotals = phyloTree->getGenusTotals();
 100
 101                                 m->mothurOut("Calculating template taxonomy tree...     "); cout.flush();
 102
 103                                 phyloTree->printTreeNodes(phyloTreeName);
 104
 105                                 m->mothurOut("DONE."); m->mothurOutEndLine();
 106
 107                                 m->mothurOut("Calculating template probabilities...     "); cout.flush();
 108
 109                                 numKmers = database->getMaxKmer() + 1;
 110
 111                                 //initialze probabilities
 112                                 wordGenusProb.resize(numKmers);
 113                                 WordPairDiffArr.resize(numKmers);
 114                         //cout << numKmers << '\t' << genusNodes.size() << endl;
 115                                 for (int j = 0; j < wordGenusProb.size(); j++) {        wordGenusProb[j].resize(genusNodes.size());             }
 116                         //cout << numKmers << '\t' << genusNodes.size() << endl;
 117                                 ofstream out;
 118                                 ofstream out2;
 119
 120                                 #ifdef USE_MPI
 121                                         int pid;
 122                                         MPI_Comm_rank(MPI_COMM_WORLD, &pid); //find out who we are
 123
 124                                         if (pid == 0) {
 125                                 #endif
 126
 127
 128                                 m->openOutputFile(probFileName, out);
 129
 130                                 //output mothur version
 131                                 out << "#" << m->getVersion() << endl;
 132
 133                                 out << numKmers << endl;
 134
 135                                 m->openOutputFile(probFileName2, out2);
 136
 137                                 //output mothur version
 138                                 out2 << "#" << m->getVersion() << endl;
 139
 140                                 #ifdef USE_MPI
 141                                         }
 142                                 #endif
 143
 144
 145                                 //for each word
 146                                 for (int i = 0; i < numKmers; i++) {
 147                                         if (m->control_pressed) {  break; }
 148
 149                                         #ifdef USE_MPI
 150                                                 MPI_Comm_rank(MPI_COMM_WORLD, &pid); //find out who we are
 151
 152                                                 if (pid == 0) {
 153                                         #endif
 154
 155                                         out << i << '\t';
 156
 157                                         #ifdef USE_MPI
 158                                                 }
 159                                         #endif
 160
 161                                         vector<int> seqsWithWordi = database->getSequencesWithKmer(i);
 162
 163                                         map<int, int> count;
 164                                         for (int k = 0; k < genusNodes.size(); k++) {  count[genusNodes[k]] = 0;  }
 165
 166                                         //for each sequence with that word
 167                                         for (int j = 0; j < seqsWithWordi.size(); j++) {
 168                                                 int temp = phyloTree->getIndex(names[seqsWithWordi[j]]);
 169                                                 count[temp]++;  //increment count of seq in this genus who have this word
 170                                         }
 171
 172                                         //probabilityInTemplate = (# of seqs with that word in template + 0.50) / (total number of seqs in template + 1);
 173                                         float probabilityInTemplate = (seqsWithWordi.size() + 0.50) / (float) (names.size() + 1);
 174                                         diffPair tempProb(log(probabilityInTemplate), 0.0);
 175                                         WordPairDiffArr[i] = tempProb;
 176
 177                                         int numNotZero = 0;
 178                                         for (int k = 0; k < genusNodes.size(); k++) {
 179                                                 //probabilityInThisTaxonomy = (# of seqs with that word in this taxonomy + probabilityInTemplate) / (total number of seqs in this taxonomy + 1);
 180
 181
 182                                                 wordGenusProb[i][k] = log((count[genusNodes[k]] + probabilityInTemplate) / (float) (genusTotals[k] + 1));
 183
 184                                                 if (count[genusNodes[k]] != 0) {
 185                                                         #ifdef USE_MPI
 186                                                                 int pid;
 187                                                                 MPI_Comm_rank(MPI_COMM_WORLD, &pid); //find out who we are
 188
 189                                                                 if (pid == 0) {
 190                                                         #endif
 191
 192                                                         out << k << '\t' << wordGenusProb[i][k] << '\t' ;
 193
 194                                                         #ifdef USE_MPI
 195                                                                 }
 196                                                         #endif
 197
 198                                                         numNotZero++;
 199                                                 }
 200                                         }
 201
 202                                         #ifdef USE_MPI
 203                                                 MPI_Comm_rank(MPI_COMM_WORLD, &pid); //find out who we are
 204
 205                                                 if (pid == 0) {
 206                                         #endif
 207
 208                                         out << endl;
 209                                         out2 << probabilityInTemplate << '\t' << numNotZero << '\t' << log(probabilityInTemplate) << endl;
 210
 211                                         #ifdef USE_MPI
 212                                                 }
 213                                         #endif
 214                                 }
 215
 216                                 #ifdef USE_MPI
 217                                         MPI_Comm_rank(MPI_COMM_WORLD, &pid); //find out who we are
 218
 219                                         if (pid == 0) {
 220                                 #endif
 221
 222                                 out.close();
 223                                 out2.close();
 224
 225                                 #ifdef USE_MPI
 226                                         }
 227                                 #endif
 228
 229                                 //read in new phylotree with less info. - its faster
 230                                 ifstream phyloTreeTest(phyloTreeName.c_str());
 231                                 delete phyloTree;
 232
 233                                 phyloTree = new PhyloTree(phyloTreeTest, phyloTreeName);
 234
 235                                 //save probabilities
 236                                 if (rdb->save) { rdb->wordGenusProb = wordGenusProb; rdb->WordPairDiffArr = WordPairDiffArr; }
 237                         }
 238                 }
 239
 240                 generateWordPairDiffArr();
 241
 242                 //save probabilities
 243                 if (rdb->save) { rdb->wordGenusProb = wordGenusProb; rdb->WordPairDiffArr = WordPairDiffArr; }
 244
 245                 m->mothurOut("DONE."); m->mothurOutEndLine();
 246                 m->mothurOut("It took " + toString(time(NULL) - start) + " seconds get probabilities. "); m->mothurOutEndLine();
 247         }
 248         catch(exception& e) {
 249                 m->errorOut(e, "Bayesian", "Bayesian");
 250                 exit(1);
 251         }
 252 }
 253 /**************************************************************************************************/
 254 Bayesian::~Bayesian() {
 255         try {
 256
 257                  delete phyloTree;
 258                  if (database != NULL) {  delete database; }
 259         }
 260         catch(exception& e) {
 261                 m->errorOut(e, "Bayesian", "~Bayesian");
 262                 exit(1);
 263         }
 264 }
 265
 266 /**************************************************************************************************/
 267 string Bayesian::getTaxonomy(Sequence* seq) {
 268         try {
 269                 string tax = "";
 270                 Kmer kmer(kmerSize);
 271                 flipped = false;
 272
 273                 //get words contained in query
 274                 //getKmerString returns a string where the index in the string is hte kmer number
 275                 //and the character at that index can be converted to be the number of times that kmer was seen
 276                 string queryKmerString = kmer.getKmerString(seq->getUnaligned());
 277
 278                 vector<int> queryKmers;
 279                 for (int i = 0; i < queryKmerString.length()-1; i++) {  // the -1 is to ignore any kmer with an N in it
 280                         if (queryKmerString[i] != '!') { //this kmer is in the query
 281                                 queryKmers.push_back(i);
 282                         }
 283                 }
 284
 285                 //if user wants to test reverse compliment and its reversed use that instead
 286                 if (flip) {
 287                         if (isReversed(queryKmers)) {
 288                                 flipped = true;
 289                                 seq->reverseComplement();
 290                                 queryKmerString = kmer.getKmerString(seq->getUnaligned());
 291                                 queryKmers.clear();
 292                                 for (int i = 0; i < queryKmerString.length()-1; i++) {  // the -1 is to ignore any kmer with an N in it
 293                                         if (queryKmerString[i] != '!') { //this kmer is in the query
 294                                                 queryKmers.push_back(i);
 295                                         }
 296                                 }
 297                         }
 298                 }
 299
 300                 if (queryKmers.size() == 0) {  m->mothurOut(seq->getName() + "is bad."); m->mothurOutEndLine(); simpleTax = "unknown;";  return "unknown;"; }
 301
 302
 303                 int index = getMostProbableTaxonomy(queryKmers);
 304
 305                 if (m->control_pressed) { return tax; }
 306
 307                 //bootstrap - to set confidenceScore
 308                 int numToSelect = queryKmers.size() / 8;
 309
 310                 tax = bootstrapResults(queryKmers, index, numToSelect);
 311
 312                 return tax;
 313         }
 314         catch(exception& e) {
 315                 m->errorOut(e, "Bayesian", "getTaxonomy");
 316                 exit(1);
 317         }
 318 }
 319 /**************************************************************************************************/
 320 string Bayesian::bootstrapResults(vector<int> kmers, int tax, int numToSelect) {
 321         try {
 322
 323                 map<int, int> confidenceScores;
 324
 325                 //initialize confidences to 0
 326                 int seqIndex = tax;
 327                 TaxNode seq = phyloTree->get(tax);
 328                 confidenceScores[tax] = 0;
 329
 330                 while (seq.level != 0) { //while you are not at the root
 331                         seqIndex = seq.parent;
 332                         confidenceScores[seqIndex] = 0;
 333                         seq = phyloTree->get(seq.parent);
 334                 }
 335
 336                 map<int, int>::iterator itBoot;
 337                 map<int, int>::iterator itBoot2;
 338                 map<int, int>::iterator itConvert;
 339
 340                 for (int i = 0; i < iters; i++) {
 341                         if (m->control_pressed) { return "control"; }
 342
 343                         vector<int> temp;
 344                         for (int j = 0; j < numToSelect; j++) {
 345                                 int index = int(rand() % kmers.size());
 346
 347                                 //add word to temp
 348                                 temp.push_back(kmers[index]);
 349                         }
 350
 351                         //get taxonomy
 352                         int newTax = getMostProbableTaxonomy(temp);
 353                         //int newTax = 1;
 354                         TaxNode taxonomyTemp = phyloTree->get(newTax);
 355
 356                         //add to confidence results
 357                         while (taxonomyTemp.level != 0) { //while you are not at the root
 358                                 itBoot2 = confidenceScores.find(newTax); //is this a classification we already have a count on
 359
 360                                 if (itBoot2 != confidenceScores.end()) { //this is a classification we need a confidence for
 361                                         (itBoot2->second)++;
 362                                 }
 363
 364                                 newTax = taxonomyTemp.parent;
 365                                 taxonomyTemp = phyloTree->get(newTax);
 366                         }
 367
 368                 }
 369
 370                 string confidenceTax = "";
 371                 simpleTax = "";
 372
 373                 int seqTaxIndex = tax;
 374                 TaxNode seqTax = phyloTree->get(tax);
 375
 376                 while (seqTax.level != 0) { //while you are not at the root
 377
 378                                 itBoot2 = confidenceScores.find(seqTaxIndex); //is this a classification we already have a count on
 379
 380                                 int confidence = 0;
 381                                 if (itBoot2 != confidenceScores.end()) { //already in confidence scores
 382                                         confidence = itBoot2->second;
 383                                 }
 384
 385                                 if (((confidence/(float)iters) * 100) >= confidenceThreshold) {
 386                                         confidenceTax = seqTax.name + "(" + toString(((confidence/(float)iters) * 100)) + ");" + confidenceTax;
 387                                         simpleTax = seqTax.name + ";" + simpleTax;
 388                                 }
 389
 390                                 seqTaxIndex = seqTax.parent;
 391                                 seqTax = phyloTree->get(seqTax.parent);
 392                 }
 393
 394                 if (confidenceTax == "") { confidenceTax = "unknown;"; simpleTax = "unknown;";  }
 395
 396                 return confidenceTax;
 397
 398         }
 399         catch(exception& e) {
 400                 m->errorOut(e, "Bayesian", "bootstrapResults");
 401                 exit(1);
 402         }
 403 }
 404 /**************************************************************************************************/
 405 int Bayesian::getMostProbableTaxonomy(vector<int> queryKmer) {
 406         try {
 407                 int indexofGenus = 0;
 408
 409                 double maxProbability = -1000000.0;
 410                 //find taxonomy with highest probability that this sequence is from it
 411
 412
 413 //              cout << genusNodes.size() << endl;
 414
 415
 416                 for (int k = 0; k < genusNodes.size(); k++) {
 417                         //for each taxonomy calc its probability
 418
 419                         double prob = 0.0000;
 420                         for (int i = 0; i < queryKmer.size(); i++) {
 421                                 prob += wordGenusProb[queryKmer[i]][k];
 422                         }
 423
 424 //                      cout << phyloTree->get(genusNodes[k]).name << '\t' << prob << endl;
 425
 426                         //is this the taxonomy with the greatest probability?
 427                         if (prob > maxProbability) {
 428                                 indexofGenus = genusNodes[k];
 429                                 maxProbability = prob;
 430                         }
 431                 }
 432
 433
 434                 return indexofGenus;
 435         }
 436         catch(exception& e) {
 437                 m->errorOut(e, "Bayesian", "getMostProbableTaxonomy");
 438                 exit(1);
 439         }
 440 }
 441 //********************************************************************************************************************
 442 //if it is more probable that the reverse compliment kmers are in the template, then we assume the sequence is reversed.
 443 bool Bayesian::isReversed(vector<int>& queryKmers){
 444         try{
 445                 bool reversed = false;
 446                 float prob = 0;
 447                 float reverseProb = 0;
 448
 449         for (int i = 0; i < queryKmers.size(); i++){
 450             int kmer = queryKmers[i];
 451             if (kmer >= 0){
 452                 prob += WordPairDiffArr[kmer].prob;
 453                                 reverseProb += WordPairDiffArr[kmer].reverseProb;
 454             }
 455         }
 456
 457         if (reverseProb > prob){ reversed = true; }
 458
 459                 return reversed;
 460         }
 461         catch(exception& e) {
 462                 m->errorOut(e, "Bayesian", "isReversed");
 463                 exit(1);
 464         }
 465 }
 466 //********************************************************************************************************************
 467 int Bayesian::generateWordPairDiffArr(){
 468         try{
 469                 Kmer kmer(kmerSize);
 470                 for (int i = 0; i < WordPairDiffArr.size(); i++) {
 471                         int reversedWord = kmer.getReverseKmerNumber(i);
 472                         WordPairDiffArr[i].reverseProb = WordPairDiffArr[reversedWord].prob;
 473                 }
 474
 475                 return 0;
 476         }catch(exception& e) {
 477                 m->errorOut(e, "Bayesian", "generateWordPairDiffArr");
 478                 exit(1);
 479         }
 480 }
 481 /*************************************************************************************************
 482 map<string, int> Bayesian::parseTaxMap(string newTax) {
 483         try{
 484
 485                 map<string, int> parsed;
 486
 487                 newTax = newTax.substr(0, newTax.length()-1);  //get rid of last ';'
 488
 489                 //parse taxonomy
 490                 string individual;
 491                 while (newTax.find_first_of(';') != -1) {
 492                         individual = newTax.substr(0,newTax.find_first_of(';'));
 493                         newTax = newTax.substr(newTax.find_first_of(';')+1, newTax.length());
 494                         parsed[individual] = 1;
 495                 }
 496
 497                 //get last one
 498                 parsed[newTax] = 1;
 499
 500                 return parsed;
 501
 502         }
 503         catch(exception& e) {
 504                 m->errorOut(e, "Bayesian", "parseTax");
 505                 exit(1);
 506         }
 507 }
 508 /**************************************************************************************************/
 509 void Bayesian::readProbFile(ifstream& in, ifstream& inNum, string inName, string inNumName) {
 510         try{
 511
 512                 #ifdef USE_MPI
 513
 514                         int pid, num, num2, processors;
 515                         vector<unsigned long long> positions;
 516                         vector<unsigned long long> positions2;
 517
 518                         MPI_Status status;
 519                         MPI_File inMPI;
 520                         MPI_File inMPI2;
 521                         MPI_Comm_rank(MPI_COMM_WORLD, &pid); //find out who we are
 522                         MPI_Comm_size(MPI_COMM_WORLD, &processors);
 523                         int tag = 2001;
 524
 525                         char inFileName[1024];
 526                         strcpy(inFileName, inNumName.c_str());
 527
 528                         char inFileName2[1024];
 529                         strcpy(inFileName2, inName.c_str());
 530
 531                         MPI_File_open(MPI_COMM_WORLD, inFileName, MPI_MODE_RDONLY, MPI_INFO_NULL, &inMPI);  //comm, filename, mode, info, filepointer
 532                         MPI_File_open(MPI_COMM_WORLD, inFileName2, MPI_MODE_RDONLY, MPI_INFO_NULL, &inMPI2);  //comm, filename, mode, info, filepointer
 533
 534                         if (pid == 0) {
 535                                 positions = m->setFilePosEachLine(inNumName, num);
 536                                 positions2 = m->setFilePosEachLine(inName, num2);
 537
 538                                 for(int i = 1; i < processors; i++) {
 539                                         MPI_Send(&num, 1, MPI_INT, i, tag, MPI_COMM_WORLD);
 540                                         MPI_Send(&positions[0], (num+1), MPI_LONG, i, tag, MPI_COMM_WORLD);
 541
 542                                         MPI_Send(&num2, 1, MPI_INT, i, tag, MPI_COMM_WORLD);
 543                                         MPI_Send(&positions2[0], (num2+1), MPI_LONG, i, tag, MPI_COMM_WORLD);
 544                                 }
 545
 546                         }else{
 547                                 MPI_Recv(&num, 1, MPI_INT, 0, tag, MPI_COMM_WORLD, &status);
 548                                 positions.resize(num+1);
 549                                 MPI_Recv(&positions[0], (num+1), MPI_LONG, 0, tag, MPI_COMM_WORLD, &status);
 550
 551                                 MPI_Recv(&num2, 1, MPI_INT, 0, tag, MPI_COMM_WORLD, &status);
 552                                 positions2.resize(num2+1);
 553                                 MPI_Recv(&positions2[0], (num2+1), MPI_LONG, 0, tag, MPI_COMM_WORLD, &status);
 554                         }
 555
 556                         //read version
 557                         int length = positions2[1] - positions2[0];
 558                         char* buf5 = new char[length];
 559
 560                         MPI_File_read_at(inMPI2, positions2[0], buf5, length, MPI_CHAR, &status);
 561                         delete buf5;
 562
 563                         //read numKmers
 564                         length = positions2[2] - positions2[1];
 565                         char* buf = new char[length];
 566
 567                         MPI_File_read_at(inMPI2, positions2[1], buf, length, MPI_CHAR, &status);
 568
 569                         string tempBuf = buf;
 570                         if (tempBuf.length() > length) { tempBuf = tempBuf.substr(0, length); }
 571                         delete buf;
 572
 573                         istringstream iss (tempBuf,istringstream::in);
 574                         iss >> numKmers;
 575
 576                         //initialze probabilities
 577                         wordGenusProb.resize(numKmers);
 578
 579                         for (int j = 0; j < wordGenusProb.size(); j++) {        wordGenusProb[j].resize(genusNodes.size());             }
 580
 581                         int kmer, name;
 582                         vector<int> numbers; numbers.resize(numKmers);
 583                         float prob;
 584                         vector<float> zeroCountProb; zeroCountProb.resize(numKmers);
 585                         WordPairDiffArr.resize(numKmers);
 586
 587                         //read version
 588                         length = positions[1] - positions[0];
 589                         char* buf6 = new char[length];
 590
 591                         MPI_File_read_at(inMPI2, positions[0], buf6, length, MPI_CHAR, &status);
 592                         delete buf6;
 593
 594                         //read file
 595                         for(int i=1;i<num;i++){
 596                                 //read next sequence
 597                                 length = positions[i+1] - positions[i];
 598                                 char* buf4 = new char[length];
 599
 600                                 MPI_File_read_at(inMPI, positions[i], buf4, length, MPI_CHAR, &status);
 601
 602                                 tempBuf = buf4;
 603                                 if (tempBuf.length() > length) { tempBuf = tempBuf.substr(0, length); }
 604                                 delete buf4;
 605
 606                                 istringstream iss (tempBuf,istringstream::in);
 607                                 float probTemp;
 608                                 iss >> zeroCountProb[i] >> numbers[i] >> probTemp;
 609                                 WordPairDiffArr[i].prob = tempProb;
 610
 611                         }
 612
 613                         MPI_File_close(&inMPI);
 614
 615                         for(int i=2;i<num2;i++){
 616                                 //read next sequence
 617                                 length = positions2[i+1] - positions2[i];
 618                                 char* buf4 = new char[length];
 619
 620                                 MPI_File_read_at(inMPI2, positions2[i], buf4, length, MPI_CHAR, &status);
 621
 622                                 tempBuf = buf4;
 623                                 if (tempBuf.length() > length) { tempBuf = tempBuf.substr(0, length); }
 624                                 delete buf4;
 625
 626                                 istringstream iss (tempBuf,istringstream::in);
 627
 628                                 iss >> kmer;
 629
 630                                 //set them all to zero value
 631                                 for (int i = 0; i < genusNodes.size(); i++) {
 632                                         wordGenusProb[kmer][i] = log(zeroCountProb[kmer] / (float) (genusTotals[i]+1));
 633                                 }
 634
 635                                 //get probs for nonzero values
 636                                 for (int i = 0; i < numbers[kmer]; i++) {
 637                                         iss >> name >> prob;
 638                                         wordGenusProb[kmer][name] = prob;
 639                                 }
 640
 641                         }
 642                         MPI_File_close(&inMPI2);
 643                         MPI_Barrier(MPI_COMM_WORLD); //make everyone wait - just in case
 644                 #else
 645                         //read version
 646                         string line = m->getline(in); m->gobble(in);
 647
 648                         in >> numKmers; m->gobble(in);
 649                         //cout << threadID << '\t' << line << '\t' << numKmers << &in << '\t' << &inNum << '\t' << genusNodes.size() << endl;
 650                         //initialze probabilities
 651                         wordGenusProb.resize(numKmers);
 652
 653                         for (int j = 0; j < wordGenusProb.size(); j++) {        wordGenusProb[j].resize(genusNodes.size());             }
 654
 655                         int kmer, name, count;  count = 0;
 656                         vector<int> num; num.resize(numKmers);
 657                         float prob;
 658                         vector<float> zeroCountProb; zeroCountProb.resize(numKmers);
 659                         WordPairDiffArr.resize(numKmers);
 660
 661                         //read version
 662                         string line2 = m->getline(inNum); m->gobble(inNum);
 663                         float probTemp;
 664                 //cout << threadID << '\t' << line2 << '\t' << this << endl;
 665                         while (inNum) {
 666                                 inNum >> zeroCountProb[count] >> num[count] >> probTemp;
 667                                 WordPairDiffArr[count].prob = probTemp;
 668                                 count++;
 669                                 m->gobble(inNum);
 670                                 //cout << threadID << '\t' << count << endl;
 671                         }
 672                         inNum.close();
 673                 //cout << threadID << '\t' << "here1 " << &wordGenusProb << '\t' << &num << endl; //
 674                 //cout << threadID << '\t' << &genusTotals << '\t' << endl;
 675                 //cout << threadID << '\t' << genusNodes.size() << endl;
 676                         while(in) {
 677                                 in >> kmer;
 678                         //cout << threadID << '\t' << kmer << endl;
 679                                 //set them all to zero value
 680                                 for (int i = 0; i < genusNodes.size(); i++) {
 681                                         wordGenusProb[kmer][i] = log(zeroCountProb[kmer] / (float) (genusTotals[i]+1));
 682                                 }
 683                         //cout << threadID << '\t' << num[kmer] << "here" << endl;
 684                                 //get probs for nonzero values
 685                                 for (int i = 0; i < num[kmer]; i++) {
 686                                         in >> name >> prob;
 687                                         wordGenusProb[kmer][name] = prob;
 688                                 }
 689
 690                                 m->gobble(in);
 691                         }
 692                         in.close();
 693                 //cout << threadID << '\t' << "here" << endl;
 694                 #endif
 695         }
 696         catch(exception& e) {
 697                 m->errorOut(e, "Bayesian", "readProbFile");
 698                 exit(1);
 699         }
 700 }
 701 /**************************************************************************************************/
 702 bool Bayesian::checkReleaseDate(ifstream& file1, ifstream& file2, ifstream& file3, ifstream& file4) {
 703         try {
 704
 705                 bool good = true;
 706
 707                 vector<string> lines;
 708                 lines.push_back(m->getline(file1));
 709                 lines.push_back(m->getline(file2));
 710                 lines.push_back(m->getline(file3));
 711                 lines.push_back(m->getline(file4));
 712
 713                 //before we added this check
 714                 if ((lines[0][0] != '#') || (lines[1][0] != '#') || (lines[2][0] != '#') || (lines[3][0] != '#')) {  good = false;  }
 715                 else {
 716                         //rip off #
 717                         for (int i = 0; i < lines.size(); i++) { lines[i] = lines[i].substr(1);  }
 718
 719                         //get mothurs current version
 720                         string version = m->getVersion();
 721
 722                         vector<string> versionVector;
 723                         m->splitAtChar(version, versionVector, '.');
 724
 725                         //check each files version
 726                         for (int i = 0; i < lines.size(); i++) {
 727                                 vector<string> linesVector;
 728                                 m->splitAtChar(lines[i], linesVector, '.');
 729
 730                                 if (versionVector.size() != linesVector.size()) { good = false; break; }
 731                                 else {
 732                                         for (int j = 0; j < versionVector.size(); j++) {
 733                                                 int num1, num2;
 734                                                 convert(versionVector[j], num1);
 735                                                 convert(linesVector[j], num2);
 736
 737                                                 //if mothurs version is newer than this files version, then we want to remake it
 738                                                 if (num1 > num2) {  good = false; break;  }
 739                                         }
 740                                 }
 741
 742                                 if (!good) { break; }
 743                         }
 744                 }
 745
 746                 if (!good) {  file1.close(); file2.close(); file3.close(); file4.close();  }
 747                 else { file1.seekg(0);  file2.seekg(0);  file3.seekg(0);  file4.seekg(0);  }
 748
 749                 return good;
 750         }
 751         catch(exception& e) {
 752                 m->errorOut(e, "Bayesian", "checkReleaseDate");
 753                 exit(1);
 754         }
 755 }
 756 /**************************************************************************************************/
 757
 758
 759
 760
 761
 762