*/
#include "chimeraslayercommand.h"
-#include "chimeraslayer.h"
#include "deconvolutecommand.h"
+#include "referencedb.h"
+#include "sequenceparser.h"
//**********************************************************************************************************************
vector<string> ChimeraSlayerCommand::setParameters(){
CommandParameter ptemplate("reference", "InputTypes", "", "", "none", "none", "none",false,true); parameters.push_back(ptemplate);
CommandParameter pfasta("fasta", "InputTypes", "", "", "none", "none", "none",false,true); parameters.push_back(pfasta);
CommandParameter pname("name", "InputTypes", "", "", "none", "none", "none",false,false); parameters.push_back(pname);
+ CommandParameter pgroup("group", "InputTypes", "", "", "none", "none", "none",false,false); parameters.push_back(pgroup);
CommandParameter pwindow("window", "Number", "", "50", "", "", "",false,false); parameters.push_back(pwindow);
CommandParameter pksize("ksize", "Number", "", "7", "", "", "",false,false); parameters.push_back(pksize);
CommandParameter pmatch("match", "Number", "", "5.0", "", "", "",false,false); parameters.push_back(pmatch);
CommandParameter pdivergence("divergence", "Number", "", "1.007", "", "", "",false,false); parameters.push_back(pdivergence);
CommandParameter pparents("parents", "Number", "", "3", "", "", "",false,false); parameters.push_back(pparents);
CommandParameter pincrement("increment", "Number", "", "5", "", "", "",false,false); parameters.push_back(pincrement);
+ CommandParameter pblastlocation("blastlocation", "String", "", "", "", "", "",false,false); parameters.push_back(pblastlocation);
CommandParameter pinputdir("inputdir", "String", "", "", "", "", "",false,false); parameters.push_back(pinputdir);
CommandParameter poutputdir("outputdir", "String", "", "", "", "", "",false,false); parameters.push_back(poutputdir);
-
+ CommandParameter psave("save", "Boolean", "", "F", "", "", "",false,false); parameters.push_back(psave);
+
vector<string> myArray;
for (int i = 0; i < parameters.size(); i++) { myArray.push_back(parameters[i].name); }
return myArray;
string helpString = "";
helpString += "The chimera.slayer command reads a fastafile and referencefile and outputs potentially chimeric sequences.\n";
helpString += "This command was modeled after the chimeraSlayer written by the Broad Institute.\n";
- helpString += "The chimera.slayer command parameters are fasta, name, template, processors, trim, ksize, window, match, mismatch, divergence. minsim, mincov, minbs, minsnp, parents, search, iters, increment, numwanted, and realign.\n";
+ helpString += "The chimera.slayer command parameters are fasta, name, template, processors, trim, ksize, window, match, mismatch, divergence. minsim, mincov, minbs, minsnp, parents, search, iters, increment, numwanted, blastlocation and realign.\n";
helpString += "The fasta parameter allows you to enter the fasta file containing your potentially chimeric sequences, and is required, unless you have a valid current fasta file. \n";
- helpString += "The name parameter allows you to provide a name file, if you are using template=self. \n";
+ helpString += "The name parameter allows you to provide a name file, if you are using reference=self. \n";
+ helpString += "The group parameter allows you to provide a group file. The group file can be used with a namesfile and reference=self. When checking sequences, only sequences from the same group as the query sequence will be used as the reference. \n";
helpString += "You may enter multiple fasta files by separating their names with dashes. ie. fasta=abrecovery.fasta-amazon.fasta \n";
helpString += "The reference parameter allows you to enter a reference file containing known non-chimeric sequences, and is required. You may also set template=self, in this case the abundant sequences will be used as potential parents. \n";
helpString += "The processors parameter allows you to specify how many processors you would like to use. The default is 1. \n";
helpString += "The minsnp parameter allows you to specify percent of SNPs to sample on each side of breakpoint for computing bootstrap support (default: 10) \n";
helpString += "The search parameter allows you to specify search method for finding the closest parent. Choices are blast, and kmer, default blast. \n";
helpString += "The realign parameter allows you to realign the query to the potential parents. Choices are true or false, default true. \n";
+ helpString += "The blastlocation parameter allows you to specify the location of your blast executable. By default mothur will look in ./blast/bin relative to mothur's executable. \n";
+ helpString += "If the save parameter is set to true the reference sequences will be saved in memory, to clear them later you can use the clear.memory command. Default=f.";
helpString += "The chimera.slayer command should be in the following format: \n";
helpString += "chimera.slayer(fasta=yourFastaFile, reference=yourTemplate, search=yourSearch) \n";
helpString += "Example: chimera.slayer(fasta=AD.align, reference=core_set_aligned.imputed.fasta, search=kmer) \n";
ChimeraSlayerCommand::ChimeraSlayerCommand(string option) {
try {
abort = false; calledHelp = false;
+ ReferenceDB* rdb = ReferenceDB::getInstance();
//allow user to run help
if(option == "help") { help(); abort = true; calledHelp = true; }
//erase from file list
fastaFileNames.erase(fastaFileNames.begin()+i);
i--;
+ }else {
+ m->setFastaFile(fastaFileNames[i]);
}
}
}
//erase from file list
nameFileNames.erase(nameFileNames.begin()+i);
i--;
+ }else {
+ m->setNameFile(nameFileNames[i]);
}
}
}
if (hasName && (nameFileNames.size() != fastaFileNames.size())) { m->mothurOut("[ERROR]: The number of namefiles does not match the number of fastafiles, please correct."); m->mothurOutEndLine(); abort=true; }
+ bool hasGroup = true;
+ groupfile = validParameter.validFile(parameters, "group", false);
+ if (groupfile == "not found") { groupfile = ""; hasGroup = false; }
+ else {
+ m->splitAtDash(groupfile, groupFileNames);
+
+ //go through files and make sure they are good, if not, then disregard them
+ for (int i = 0; i < groupFileNames.size(); i++) {
+
+ bool ignore = false;
+ if (groupFileNames[i] == "current") {
+ groupFileNames[i] = m->getGroupFile();
+ if (groupFileNames[i] != "") { m->mothurOut("Using " + groupFileNames[i] + " as input file for the group parameter where you had given current."); m->mothurOutEndLine(); }
+ else {
+ m->mothurOut("You have no current namefile, ignoring current."); m->mothurOutEndLine(); ignore=true;
+ //erase from file list
+ groupFileNames.erase(groupFileNames.begin()+i);
+ i--;
+ }
+ }
+
+ if (!ignore) {
+
+ if (inputDir != "") {
+ string path = m->hasPath(groupFileNames[i]);
+ //if the user has not given a path then, add inputdir. else leave path alone.
+ if (path == "") { groupFileNames[i] = inputDir + groupFileNames[i]; }
+ }
+
+ int ableToOpen;
+ ifstream in;
+
+ ableToOpen = m->openInputFile(groupFileNames[i], in, "noerror");
+
+ //if you can't open it, try default location
+ if (ableToOpen == 1) {
+ if (m->getDefaultPath() != "") { //default path is set
+ string tryPath = m->getDefaultPath() + m->getSimpleName(groupFileNames[i]);
+ m->mothurOut("Unable to open " + groupFileNames[i] + ". Trying default " + tryPath); m->mothurOutEndLine();
+ ifstream in2;
+ ableToOpen = m->openInputFile(tryPath, in2, "noerror");
+ in2.close();
+ groupFileNames[i] = tryPath;
+ }
+ }
+
+ if (ableToOpen == 1) {
+ if (m->getOutputDir() != "") { //default path is set
+ string tryPath = m->getOutputDir() + m->getSimpleName(groupFileNames[i]);
+ m->mothurOut("Unable to open " + groupFileNames[i] + ". Trying output directory " + tryPath); m->mothurOutEndLine();
+ ifstream in2;
+ ableToOpen = m->openInputFile(tryPath, in2, "noerror");
+ in2.close();
+ groupFileNames[i] = tryPath;
+ }
+ }
+
+ in.close();
+
+ if (ableToOpen == 1) {
+ m->mothurOut("Unable to open " + groupFileNames[i] + ". It will be disregarded."); m->mothurOutEndLine();
+ //erase from file list
+ groupFileNames.erase(groupFileNames.begin()+i);
+ i--;
+ }else {
+ m->setGroupFile(groupFileNames[i]);
+ }
+ }
+ }
+
+ //make sure there is at least one valid file left
+ if (groupFileNames.size() == 0) { m->mothurOut("[ERROR]: no valid group files."); m->mothurOutEndLine(); abort = true; }
+ }
+
+ if (hasGroup && (groupFileNames.size() != fastaFileNames.size())) { m->mothurOut("[ERROR]: The number of groupfiles does not match the number of fastafiles, please correct."); m->mothurOutEndLine(); abort=true; }
+
+
//if the user changes the output directory command factory will send this info to us in the output parameter
outputDir = validParameter.validFile(parameters, "outputdir", false); if (outputDir == "not found"){ outputDir = ""; }
+ string temp = validParameter.validFile(parameters, "processors", false); if (temp == "not found"){ temp = m->getProcessors(); }
+ m->setProcessors(temp);
+ convert(temp, processors);
+
+ temp = validParameter.validFile(parameters, "save", false); if (temp == "not found"){ temp = "f"; }
+ save = m->isTrue(temp);
+ rdb->save = save;
+ if (save) { //clear out old references
+ rdb->clearMemory();
+ }
string path;
it = parameters.find("reference");
//user has given a template file
if(it != parameters.end()){
- if (it->second == "self") { templatefile = "self"; }
+ if (it->second == "self") {
+ templatefile = "self";
+ if (save) {
+ m->mothurOut("[WARNING]: You can't save reference=self, ignoring save.");
+ m->mothurOutEndLine();
+ save = false;
+ }
+ }
else {
path = m->hasPath(it->second);
//if the user has not given a path then, add inputdir. else leave path alone.
templatefile = validParameter.validFile(parameters, "reference", true);
if (templatefile == "not open") { abort = true; }
- else if (templatefile == "not found") { templatefile = ""; m->mothurOut("reference is a required parameter for the chimera.slayer command, unless and namefile is given."); m->mothurOutEndLine(); abort = true; }
+ else if (templatefile == "not found") { //check for saved reference sequences
+ if (rdb->referenceSeqs.size() != 0) {
+ templatefile = "saved";
+ }else {
+ m->mothurOut("[ERROR]: You don't have any saved reference sequences and the reference parameter is a required.");
+ m->mothurOutEndLine();
+ abort = true;
+ }
+ }else { if (save) { rdb->setSavedReference(templatefile); } }
+ }
+ }else if (hasName) { templatefile = "self";
+ if (save) {
+ m->mothurOut("[WARNING]: You can't save reference=self, ignoring save.");
+ m->mothurOutEndLine();
+ save = false;
}
- }else if (hasName) { templatefile = "self"; }
- else { templatefile = ""; m->mothurOut("reference is a required parameter for the chimera.slayer command."); m->mothurOutEndLine(); abort = true; }
+ }
+ else {
+ if (rdb->referenceSeqs.size() != 0) {
+ templatefile = "saved";
+ }else {
+ m->mothurOut("[ERROR]: You don't have any saved reference sequences and the reference parameter is a required.");
+ m->mothurOutEndLine();
+ templatefile = ""; abort = true;
+ }
+ }
+
- string temp = validParameter.validFile(parameters, "processors", false); if (temp == "not found"){ temp = m->getProcessors(); }
- m->setProcessors(temp);
- convert(temp, processors);
temp = validParameter.validFile(parameters, "ksize", false); if (temp == "not found") { temp = "7"; }
convert(temp, ksize);
temp = validParameter.validFile(parameters, "numwanted", false); if (temp == "not found") { temp = "15"; }
convert(temp, numwanted);
+
+ blastlocation = validParameter.validFile(parameters, "blastlocation", false);
+ if (blastlocation == "not found") { blastlocation = ""; }
+ else {
+ //add / to name if needed
+ string lastChar = blastlocation.substr(blastlocation.length()-1);
+#if defined (__APPLE__) || (__MACH__) || (linux) || (__linux)
+ if (lastChar != "/") { blastlocation += "/"; }
+#else
+ if (lastChar != "\\") { blastlocation += "\\"; }
+#endif
+ blastlocation = m->getFullPathName(blastlocation);
+ string formatdbCommand = "";
+#if defined (__APPLE__) || (__MACH__) || (linux) || (__linux)
+ formatdbCommand = blastlocation + "formatdb";
+#else
+ formatdbCommand = blastlocation + "formatdb.exe";
+#endif
+
+ //test to make sure formatdb exists
+ ifstream in;
+ formatdbCommand = m->getFullPathName(formatdbCommand);
+ int ableToOpen = m->openInputFile(formatdbCommand, in, "no error"); in.close();
+ if(ableToOpen == 1) { m->mothurOut("[ERROR]: " + formatdbCommand + " file does not exist. mothur requires formatdb.exe to run chimera.slayer."); m->mothurOutEndLine(); abort = true; }
+
+ string blastCommand = "";
+#if defined (__APPLE__) || (__MACH__) || (linux) || (__linux)
+ blastCommand = blastlocation + "megablast";
+#else
+ blastCommand = blastlocation + "megablast.exe";
+#endif
+ //test to make sure formatdb exists
+ ifstream in2;
+ blastCommand = m->getFullPathName(blastCommand);
+ ableToOpen = m->openInputFile(blastCommand, in2, "no error"); in2.close();
+ if(ableToOpen == 1) { m->mothurOut("[ERROR]: " + blastCommand + " file does not exist. mothur requires blastall.exe to run chimera.slayer."); m->mothurOutEndLine(); abort = true; }
+ }
if ((search != "blast") && (search != "kmer")) { m->mothurOut(search + " is not a valid search."); m->mothurOutEndLine(); abort = true; }
+
+ if (hasName && (templatefile != "self")) { m->mothurOut("You have provided a namefile and the reference parameter is not set to self. I am not sure what reference you are trying to use, aborting."); m->mothurOutEndLine(); abort=true; }
+ if (hasGroup && (templatefile != "self")) { m->mothurOut("You have provided a group file and the reference parameter is not set to self. I am not sure what reference you are trying to use, aborting."); m->mothurOutEndLine(); abort=true; }
+
}
}
catch(exception& e) {
int ChimeraSlayerCommand::execute(){
try{
if (abort == true) { if (calledHelp) { return 0; } return 2; }
-
+
for (int s = 0; s < fastaFileNames.size(); s++) {
m->mothurOut("Checking sequences from " + fastaFileNames[s] + " ..." ); m->mothurOutEndLine();
int start = time(NULL);
- if (templatefile != "self") { //you want to run slayer with a refernce template
- chimera = new ChimeraSlayer(fastaFileNames[s], templatefile, trim, search, ksize, match, mismatch, window, divR, minSimilarity, minCoverage, minBS, minSNP, parents, iters, increment, numwanted, realign);
- }else {
+ //you provided a groupfile
+ string groupFile = "";
+ if (groupFileNames.size() != 0) { groupFile = groupFileNames[s]; }
+
+ //maps a filename to priority map.
+ //if no groupfile this is fastafileNames[s] -> prioirity
+ //if groupfile then this is each groups seqs -> priority
+ map<string, map<string, int> > fileToPriority;
+ map<string, map<string, int> >::iterator itFile;
+ map<string, string> fileGroup;
+ map<string, int> priority;
+ fileToPriority[fastaFileNames[s]] = priority; //default
+ fileGroup[fastaFileNames[s]] = "noGroup";
+ SequenceParser* parser = NULL;
+ int totalSeqs = 0;
+ int totalChimeras = 0;
+
+ if ((templatefile == "self") && (groupFile == "")) {
+ fileGroup.clear();
+ fileToPriority.clear();
if (processors != 1) { m->mothurOut("When using template=self, mothur can only use 1 processor, continuing."); m->mothurOutEndLine(); processors = 1; }
string nameFile = "";
if (nameFileNames.size() != 0) { //you provided a namefile and we don't need to create one
nameFile = nameFileNames[s];
- }else {
- m->mothurOutEndLine(); m->mothurOut("No namesfile given, running unique.seqs command to generate one."); m->mothurOutEndLine(); m->mothurOutEndLine();
-
- //use unique.seqs to create new name and fastafile
- string inputString = "fasta=" + fastaFileNames[s];
- m->mothurOut("/******************************************/"); m->mothurOutEndLine();
- m->mothurOut("Running command: unique.seqs(" + inputString + ")"); m->mothurOutEndLine();
-
- Command* uniqueCommand = new DeconvoluteCommand(inputString);
- uniqueCommand->execute();
-
- map<string, vector<string> > filenames = uniqueCommand->getOutputFiles();
-
- delete uniqueCommand;
-
- m->mothurOut("/******************************************/"); m->mothurOutEndLine();
-
- nameFile = filenames["name"][0];
- fastaFileNames[s] = filenames["fasta"][0];
- }
+ }else { nameFile = getNamesFile(fastaFileNames[s]); }
//sort fastafile by abundance, returns new sorted fastafile name
m->mothurOut("Sorting fastafile according to abundance..."); cout.flush();
- map<string, int> priority = sortFastaFile(fastaFileNames[s], nameFile);
+ priority = sortFastaFile(fastaFileNames[s], nameFile);
m->mothurOut("Done."); m->mothurOutEndLine();
- if (m->control_pressed) { for (int j = 0; j < outputNames.size(); j++) { remove(outputNames[j].c_str()); } return 0; }
-
- chimera = new ChimeraSlayer(fastaFileNames[s], templatefile, trim, priority, search, ksize, match, mismatch, window, divR, minSimilarity, minCoverage, minBS, minSNP, parents, iters, increment, numwanted, realign);
- }
+ fileToPriority[fastaFileNames[s]] = priority;
+ if (m->control_pressed) { for (int j = 0; j < outputNames.size(); j++) { m->mothurRemove(outputNames[j]); } return 0; }
+ }else if ((templatefile == "self") && (groupFile != "")) {
+ fileGroup.clear();
+ fileToPriority.clear();
+ string nameFile = "";
+ if (nameFileNames.size() != 0) { //you provided a namefile and we don't need to create one
+ nameFile = nameFileNames[s];
+ }else { nameFile = getNamesFile(fastaFileNames[s]); }
+
+ //Parse sequences by group
+ parser = new SequenceParser(groupFile, fastaFileNames[s], nameFile);
+ vector<string> groups = parser->getNamesOfGroups();
+ for (int i = 0; i < groups.size(); i++) {
+ vector<Sequence> thisGroupsSeqs = parser->getSeqs(groups[i]);
+ map<string, string> thisGroupsMap = parser->getNameMap(groups[i]);
+ string newFastaFile = outputDir + m->getRootName(m->getSimpleName(fastaFileNames[s])) + groups[i] + "-sortedTemp.fasta";
+ priority = sortFastaFile(thisGroupsSeqs, thisGroupsMap, newFastaFile);
+ fileToPriority[newFastaFile] = priority;
+ fileGroup[newFastaFile] = groups[i];
+ }
+ }
+
if (outputDir == "") { outputDir = m->hasPath(fastaFileNames[s]); }//if user entered a file with a path then preserve it
string outputFileName = outputDir + m->getRootName(m->getSimpleName(fastaFileNames[s])) + "slayer.chimera";
string accnosFileName = outputDir + m->getRootName(m->getSimpleName(fastaFileNames[s])) + "slayer.accnos";
string trimFastaFileName = outputDir + m->getRootName(m->getSimpleName(fastaFileNames[s])) + "slayer.fasta";
- if (m->control_pressed) { delete chimera; for (int j = 0; j < outputNames.size(); j++) { remove(outputNames[j].c_str()); } return 0; }
+ //clears files
+ ofstream out, out1, out2;
+ m->openOutputFile(outputFileName, out); out.close();
+ m->openOutputFile(accnosFileName, out1); out1.close();
+ if (trim) { m->openOutputFile(trimFastaFileName, out2); out2.close(); }
+ outputNames.push_back(outputFileName); outputTypes["chimera"].push_back(outputFileName);
+ outputNames.push_back(accnosFileName); outputTypes["accnos"].push_back(accnosFileName);
+ if (trim) { outputNames.push_back(trimFastaFileName); outputTypes["fasta"].push_back(trimFastaFileName); }
- if (chimera->getUnaligned()) {
- m->mothurOut("Your template sequences are different lengths, please correct."); m->mothurOutEndLine();
- delete chimera;
- return 0;
- }
- templateSeqsLength = chimera->getLength();
- #ifdef USE_MPI
- int pid, numSeqsPerProcessor;
- int tag = 2001;
- vector<unsigned long int> MPIPos;
+ for (itFile = fileToPriority.begin(); itFile != fileToPriority.end(); itFile++) {
- MPI_Status status;
- MPI_Comm_rank(MPI_COMM_WORLD, &pid); //find out who we are
- MPI_Comm_size(MPI_COMM_WORLD, &processors);
-
- MPI_File inMPI;
- MPI_File outMPI;
- MPI_File outMPIAccnos;
- MPI_File outMPIFasta;
+ string thisFastaName = itFile->first;
+ map<string, int> thisPriority = itFile->second;
- int outMode=MPI_MODE_CREATE|MPI_MODE_WRONLY;
- int inMode=MPI_MODE_RDONLY;
+ //this is true when you have parsed by groups
+ if (fileToPriority.size() > 1) { m->mothurOutEndLine(); m->mothurOut("Checking sequences from group: " + fileGroup[thisFastaName] + "."); m->mothurOutEndLine(); }
- char outFilename[1024];
- strcpy(outFilename, outputFileName.c_str());
+ string thisoutputFileName = outputDir + m->getRootName(m->getSimpleName(thisFastaName)) + fileGroup[thisFastaName] + "slayer.chimera";
+ string thisaccnosFileName = outputDir + m->getRootName(m->getSimpleName(thisFastaName)) + fileGroup[thisFastaName] + "slayer.accnos";
+ string thistrimFastaFileName = outputDir + m->getRootName(m->getSimpleName(thisFastaName)) + fileGroup[thisFastaName] + "slayer.fasta";
- char outAccnosFilename[1024];
- strcpy(outAccnosFilename, accnosFileName.c_str());
-
- char outFastaFilename[1024];
- strcpy(outFastaFilename, trimFastaFileName.c_str());
+ //create chimera here if you are mac or linux because fork will copy for you. Create in create processes instead if you are windows.
+ if (processors == 1) { templateSeqsLength = setupChimera(thisFastaName, thisPriority); }
+ else {
+ #if defined (__APPLE__) || (__MACH__) || (linux) || (__linux) || USE_MPI
+ templateSeqsLength = setupChimera(thisFastaName, thisPriority);
+ #endif
+ }
- char inFileName[1024];
- strcpy(inFileName, fastaFileNames[s].c_str());
-
- MPI_File_open(MPI_COMM_WORLD, inFileName, inMode, MPI_INFO_NULL, &inMPI); //comm, filename, mode, info, filepointer
- MPI_File_open(MPI_COMM_WORLD, outFilename, outMode, MPI_INFO_NULL, &outMPI);
- MPI_File_open(MPI_COMM_WORLD, outAccnosFilename, outMode, MPI_INFO_NULL, &outMPIAccnos);
- if (trim) { MPI_File_open(MPI_COMM_WORLD, outFastaFilename, outMode, MPI_INFO_NULL, &outMPIFasta); }
-
- if (m->control_pressed) { outputTypes.clear(); MPI_File_close(&inMPI); MPI_File_close(&outMPI); if (trim) { MPI_File_close(&outMPIFasta); } MPI_File_close(&outMPIAccnos); for (int j = 0; j < outputNames.size(); j++) { remove(outputNames[j].c_str()); } delete chimera; return 0; }
-
- if (pid == 0) { //you are the root process
- m->mothurOutEndLine();
- m->mothurOut("Only reporting sequence supported by " + toString(minBS) + "% of bootstrapped results.");
- m->mothurOutEndLine();
-
- string outTemp = "Name\tLeftParent\tRightParent\tDivQLAQRB\tPerIDQLAQRB\tBootStrapA\tDivQLBQRA\tPerIDQLBQRA\tBootStrapB\tFlag\tLeftWindow\tRightWindow\n";
-
- //print header
- int length = outTemp.length();
- char* buf2 = new char[length];
- memcpy(buf2, outTemp.c_str(), length);
-
- MPI_File_write_shared(outMPI, buf2, length, MPI_CHAR, &status);
- delete buf2;
-
- MPIPos = m->setFilePosFasta(fastaFileNames[s], numSeqs); //fills MPIPos, returns numSeqs
+ if (m->control_pressed) { if (parser != NULL) { delete parser; } for (int j = 0; j < outputNames.size(); j++) { m->mothurRemove(outputNames[j]); } return 0; }
+
+ #ifdef USE_MPI
+ MPIExecute(thisFastaName, thisoutputFileName, thisaccnosFileName, thistrimFastaFileName);
+ if (m->control_pressed) { outputTypes.clear(); for (int j = 0; j < outputNames.size(); j++) { m->mothurRemove(outputNames[j]); } return 0; }
+ #else
+ //break up file
+ vector<unsigned long long> positions;
+ #if defined (__APPLE__) || (__MACH__) || (linux) || (__linux)
+ positions = m->divideFile(thisFastaName, processors);
+ for (int i = 0; i < (positions.size()-1); i++) { lines.push_back(new linePair(positions[i], positions[(i+1)])); }
+ #else
+ if (processors == 1) { lines.push_back(new linePair(0, 1000)); }
+ else {
+ positions = m->setFilePosFasta(thisFastaName, numSeqs);
- if (templatefile != "self") { //if template=self we can only use 1 processor
- //send file positions to all processes
- for(int i = 1; i < processors; i++) {
- MPI_Send(&numSeqs, 1, MPI_INT, i, tag, MPI_COMM_WORLD);
- MPI_Send(&MPIPos[0], (numSeqs+1), MPI_LONG, i, tag, MPI_COMM_WORLD);
+ //figure out how many sequences you have to process
+ int numSeqsPerProcessor = numSeqs / processors;
+ for (int i = 0; i < processors; i++) {
+ int startIndex = i * numSeqsPerProcessor;
+ if(i == (processors - 1)){ numSeqsPerProcessor = numSeqs - i * numSeqsPerProcessor; }
+ lines.push_back(new linePair(positions[startIndex], numSeqsPerProcessor));
}
}
- //figure out how many sequences you have to align
- numSeqsPerProcessor = numSeqs / processors;
- int startIndex = pid * numSeqsPerProcessor;
- if(pid == (processors - 1)){ numSeqsPerProcessor = numSeqs - pid * numSeqsPerProcessor; }
-
- if (templatefile == "self") { //if template=self we can only use 1 processor
- startIndex = 0;
- numSeqsPerProcessor = numSeqs;
- }
-
- //do your part
- driverMPI(startIndex, numSeqsPerProcessor, inMPI, outMPI, outMPIAccnos, outMPIFasta, MPIPos);
-
- if (m->control_pressed) { outputTypes.clear(); MPI_File_close(&inMPI); MPI_File_close(&outMPI); if (trim) { MPI_File_close(&outMPIFasta); } MPI_File_close(&outMPIAccnos); for (int j = 0; j < outputNames.size(); j++) { remove(outputNames[j].c_str()); } remove(outputFileName.c_str()); remove(accnosFileName.c_str()); delete chimera; return 0; }
-
- }else{ //you are a child process
- if (templatefile != "self") { //if template=self we can only use 1 processor
- MPI_Recv(&numSeqs, 1, MPI_INT, 0, tag, MPI_COMM_WORLD, &status);
- MPIPos.resize(numSeqs+1);
- MPI_Recv(&MPIPos[0], (numSeqs+1), MPI_LONG, 0, tag, MPI_COMM_WORLD, &status);
-
- //figure out how many sequences you have to align
- numSeqsPerProcessor = numSeqs / processors;
- int startIndex = pid * numSeqsPerProcessor;
- if(pid == (processors - 1)){ numSeqsPerProcessor = numSeqs - pid * numSeqsPerProcessor; }
-
- //do your part
- driverMPI(startIndex, numSeqsPerProcessor, inMPI, outMPI, outMPIAccnos, outMPIFasta, MPIPos);
-
- if (m->control_pressed) { outputTypes.clear(); MPI_File_close(&inMPI); MPI_File_close(&outMPI); if (trim) { MPI_File_close(&outMPIFasta); } MPI_File_close(&outMPIAccnos); for (int j = 0; j < outputNames.size(); j++) { remove(outputNames[j].c_str()); } delete chimera; return 0; }
+ #endif
- }
- }
-
- //close files
- MPI_File_close(&inMPI);
- MPI_File_close(&outMPI);
- MPI_File_close(&outMPIAccnos);
- if (trim) { MPI_File_close(&outMPIFasta); }
- MPI_Barrier(MPI_COMM_WORLD); //make everyone wait - just in case
-
- #else
- ofstream outHeader;
- string tempHeader = outputDir + m->getRootName(m->getSimpleName(fastaFileNames[s])) + "slayer.chimeras.tempHeader";
- m->openOutputFile(tempHeader, outHeader);
-
- chimera->printHeader(outHeader);
- outHeader.close();
-
- vector<unsigned long int> positions = m->divideFile(fastaFileNames[s], processors);
-
- for (int i = 0; i < (positions.size()-1); i++) {
- lines.push_back(new linePair(positions[i], positions[(i+1)]));
- }
-
- //break up file
- #if defined (__APPLE__) || (__MACH__) || (linux) || (__linux)
if(processors == 1){
- numSeqs = driver(lines[0], outputFileName, fastaFileNames[s], accnosFileName, trimFastaFileName);
-
- if (m->control_pressed) { outputTypes.clear(); if (trim) { remove(trimFastaFileName.c_str()); } remove(outputFileName.c_str()); remove(tempHeader.c_str()); remove(accnosFileName.c_str()); for (int j = 0; j < outputNames.size(); j++) { remove(outputNames[j].c_str()); } for (int i = 0; i < lines.size(); i++) { delete lines[i]; } lines.clear(); delete chimera; return 0; }
+ numSeqs = driver(lines[0], thisoutputFileName, thisFastaName, thisaccnosFileName, thistrimFastaFileName);
- }else{
- processIDS.resize(0);
+ int numNoParents = chimera->getNumNoParents();
+ if (numNoParents == numSeqs) { m->mothurOut("[WARNING]: megablast returned 0 potential parents for all your sequences. This could be due to formatdb.exe not being setup properly, please check formatdb.log for errors."); m->mothurOutEndLine(); }
- numSeqs = createProcesses(outputFileName, fastaFileNames[s], accnosFileName, trimFastaFileName);
+ }else{ numSeqs = createProcesses(thisoutputFileName, thisFastaName, thisaccnosFileName, thistrimFastaFileName); }
- rename((outputFileName + toString(processIDS[0]) + ".temp").c_str(), outputFileName.c_str());
- rename((accnosFileName + toString(processIDS[0]) + ".temp").c_str(), accnosFileName.c_str());
- if (trim) { rename((trimFastaFileName + toString(processIDS[0]) + ".temp").c_str(), trimFastaFileName.c_str()); }
-
- //append output files
- for(int i=1;i<processors;i++){
- m->appendFiles((outputFileName + toString(processIDS[i]) + ".temp"), outputFileName);
- remove((outputFileName + toString(processIDS[i]) + ".temp").c_str());
- }
-
- //append output files
- for(int i=1;i<processors;i++){
- m->appendFiles((accnosFileName + toString(processIDS[i]) + ".temp"), accnosFileName);
- remove((accnosFileName + toString(processIDS[i]) + ".temp").c_str());
- }
-
- if (trim) {
- for(int i=1;i<processors;i++){
- m->appendFiles((trimFastaFileName + toString(processIDS[i]) + ".temp"), trimFastaFileName);
- remove((trimFastaFileName + toString(processIDS[i]) + ".temp").c_str());
- }
- }
-
- if (m->control_pressed) { outputTypes.clear(); if (trim) { remove(trimFastaFileName.c_str()); } remove(outputFileName.c_str()); remove(accnosFileName.c_str()); for (int j = 0; j < outputNames.size(); j++) { remove(outputNames[j].c_str()); } for (int i = 0; i < lines.size(); i++) { delete lines[i]; } lines.clear(); delete chimera; return 0; }
- }
-
- #else
- numSeqs = driver(lines[0], outputFileName, fastaFileNames[s], accnosFileName, trimFastaFileName);
+ if (m->control_pressed) { if (parser != NULL) { delete parser; } outputTypes.clear(); if (trim) { m->mothurRemove(trimFastaFileName); } m->mothurRemove(outputFileName); m->mothurRemove(accnosFileName); for (int j = 0; j < outputNames.size(); j++) { m->mothurRemove(outputNames[j]); } for (int i = 0; i < lines.size(); i++) { delete lines[i]; } lines.clear(); delete chimera; return 0; }
- if (m->control_pressed) { outputTypes.clear(); if (trim) { remove(trimFastaFileName.c_str()); } remove(outputFileName.c_str()); remove(tempHeader.c_str()); remove(accnosFileName.c_str()); for (int j = 0; j < outputNames.size(); j++) { remove(outputNames[j].c_str()); } for (int i = 0; i < lines.size(); i++) { delete lines[i]; } lines.clear(); delete chimera; return 0; }
+ #endif
+ #if defined (__APPLE__) || (__MACH__) || (linux) || (__linux) || USE_MPI
+ delete chimera;
#endif
+
+ for (int i = 0; i < lines.size(); i++) { delete lines[i]; } lines.clear();
+
+ //append files
+ m->appendFiles(thisoutputFileName, outputFileName); m->mothurRemove(thisoutputFileName);
+ totalChimeras = m->appendFiles(thisaccnosFileName, accnosFileName); m->mothurRemove(thisaccnosFileName);
+ if (trim) { m->appendFiles(thistrimFastaFileName, trimFastaFileName); m->mothurRemove(thistrimFastaFileName); }
+
+ totalSeqs += numSeqs;
+ }
- m->appendFiles(outputFileName, tempHeader);
-
- remove(outputFileName.c_str());
- rename(tempHeader.c_str(), outputFileName.c_str());
-
- #endif
- delete chimera;
-
-
- for (int i = 0; i < lines.size(); i++) { delete lines[i]; } lines.clear();
+ if (fileToPriority.size() > 1) { totalChimeras = deconvoluteResults(parser, outputFileName, accnosFileName, trimFastaFileName); }
- outputNames.push_back(outputFileName); outputTypes["chimera"].push_back(outputFileName);
- outputNames.push_back(accnosFileName); outputTypes["accnos"].push_back(accnosFileName);
- if (trim) { outputNames.push_back(trimFastaFileName); outputTypes["fasta"].push_back(trimFastaFileName); }
+ if (parser != NULL) { delete parser; }
- m->mothurOutEndLine(); m->mothurOut("It took " + toString(time(NULL) - start) + " secs to check " + toString(numSeqs) + " sequences."); m->mothurOutEndLine();
+ m->mothurOutEndLine(); m->mothurOut(toString(totalChimeras) + " chimera found."); m->mothurOutEndLine(); m->mothurOut("It took " + toString(time(NULL) - start) + " secs to check " + toString(totalSeqs) + " sequences."); m->mothurOutEndLine();
}
//set accnos file as new current accnosfile
}
}
//**********************************************************************************************************************
+int ChimeraSlayerCommand::MPIExecute(string inputFile, string outputFileName, string accnosFileName, string trimFastaFileName){
+ try {
+
+#ifdef USE_MPI
+ int pid, numSeqsPerProcessor;
+ int tag = 2001;
+ vector<unsigned long long> MPIPos;
+
+ MPI_Status status;
+ MPI_Comm_rank(MPI_COMM_WORLD, &pid); //find out who we are
+ MPI_Comm_size(MPI_COMM_WORLD, &processors);
+
+ MPI_File inMPI;
+ MPI_File outMPI;
+ MPI_File outMPIAccnos;
+ MPI_File outMPIFasta;
+
+ int outMode=MPI_MODE_CREATE|MPI_MODE_WRONLY;
+ int inMode=MPI_MODE_RDONLY;
+
+ char outFilename[1024];
+ strcpy(outFilename, outputFileName.c_str());
+
+ char outAccnosFilename[1024];
+ strcpy(outAccnosFilename, accnosFileName.c_str());
+
+ char outFastaFilename[1024];
+ strcpy(outFastaFilename, trimFastaFileName.c_str());
+
+ char inFileName[1024];
+ strcpy(inFileName, inputFile.c_str());
+
+ MPI_File_open(MPI_COMM_WORLD, inFileName, inMode, MPI_INFO_NULL, &inMPI); //comm, filename, mode, info, filepointer
+ MPI_File_open(MPI_COMM_WORLD, outFilename, outMode, MPI_INFO_NULL, &outMPI);
+ MPI_File_open(MPI_COMM_WORLD, outAccnosFilename, outMode, MPI_INFO_NULL, &outMPIAccnos);
+ if (trim) { MPI_File_open(MPI_COMM_WORLD, outFastaFilename, outMode, MPI_INFO_NULL, &outMPIFasta); }
+
+ if (m->control_pressed) { MPI_File_close(&inMPI); MPI_File_close(&outMPI); if (trim) { MPI_File_close(&outMPIFasta); } MPI_File_close(&outMPIAccnos); delete chimera; return 0; }
+
+ if (pid == 0) { //you are the root process
+ m->mothurOutEndLine();
+ m->mothurOut("Only reporting sequence supported by " + toString(minBS) + "% of bootstrapped results.");
+ m->mothurOutEndLine();
+
+ string outTemp = "Name\tLeftParent\tRightParent\tDivQLAQRB\tPerIDQLAQRB\tBootStrapA\tDivQLBQRA\tPerIDQLBQRA\tBootStrapB\tFlag\tLeftWindow\tRightWindow\n";
+
+ //print header
+ int length = outTemp.length();
+ char* buf2 = new char[length];
+ memcpy(buf2, outTemp.c_str(), length);
+
+ MPI_File_write_shared(outMPI, buf2, length, MPI_CHAR, &status);
+ delete buf2;
+
+ MPIPos = m->setFilePosFasta(inputFile, numSeqs); //fills MPIPos, returns numSeqs
+
+ if (templatefile != "self") { //if template=self we can only use 1 processor
+ //send file positions to all processes
+ for(int i = 1; i < processors; i++) {
+ MPI_Send(&numSeqs, 1, MPI_INT, i, tag, MPI_COMM_WORLD);
+ MPI_Send(&MPIPos[0], (numSeqs+1), MPI_LONG, i, tag, MPI_COMM_WORLD);
+ }
+ }
+ //figure out how many sequences you have to align
+ numSeqsPerProcessor = numSeqs / processors;
+ int startIndex = pid * numSeqsPerProcessor;
+ if(pid == (processors - 1)){ numSeqsPerProcessor = numSeqs - pid * numSeqsPerProcessor; }
+
+ if (templatefile == "self") { //if template=self we can only use 1 processor
+ startIndex = 0;
+ numSeqsPerProcessor = numSeqs;
+ }
+
+ //do your part
+ driverMPI(startIndex, numSeqsPerProcessor, inMPI, outMPI, outMPIAccnos, outMPIFasta, MPIPos);
+
+ int numNoParents = chimera->getNumNoParents();
+ int temp;
+ for(int i = 1; i < processors; i++) {
+ MPI_Recv(&temp, 1, MPI_INT, 1, tag, MPI_COMM_WORLD, &status);
+ numNoParents += temp;
+ }
+
+
+ if (numSeqs == numNoParents) { m->mothurOut("[WARNING]: megablast returned 0 potential parents for all your sequences. This could be due to formatdb.exe not being setup properly, please check formatdb.log for errors."); m->mothurOutEndLine(); }
+
+ if (m->control_pressed) { MPI_File_close(&inMPI); MPI_File_close(&outMPI); if (trim) { MPI_File_close(&outMPIFasta); } MPI_File_close(&outMPIAccnos); delete chimera; return 0; }
+
+ }else{ //you are a child process
+ if (templatefile != "self") { //if template=self we can only use 1 processor
+ MPI_Recv(&numSeqs, 1, MPI_INT, 0, tag, MPI_COMM_WORLD, &status);
+ MPIPos.resize(numSeqs+1);
+ MPI_Recv(&MPIPos[0], (numSeqs+1), MPI_LONG, 0, tag, MPI_COMM_WORLD, &status);
+
+ //figure out how many sequences you have to align
+ numSeqsPerProcessor = numSeqs / processors;
+ int startIndex = pid * numSeqsPerProcessor;
+ if(pid == (processors - 1)){ numSeqsPerProcessor = numSeqs - pid * numSeqsPerProcessor; }
+
+ //do your part
+ driverMPI(startIndex, numSeqsPerProcessor, inMPI, outMPI, outMPIAccnos, outMPIFasta, MPIPos);
+
+ int numNoParents = chimera->getNumNoParents();
+ MPI_Send(&numNoParents, 1, MPI_INT, 0, tag, MPI_COMM_WORLD);
+
+ if (m->control_pressed) { MPI_File_close(&inMPI); MPI_File_close(&outMPI); if (trim) { MPI_File_close(&outMPIFasta); } MPI_File_close(&outMPIAccnos); delete chimera; return 0; }
+
+ }
+ }
+
+ //close files
+ MPI_File_close(&inMPI);
+ MPI_File_close(&outMPI);
+ MPI_File_close(&outMPIAccnos);
+ if (trim) { MPI_File_close(&outMPIFasta); }
+ MPI_Barrier(MPI_COMM_WORLD); //make everyone wait - just in case
+
+
+#endif
+ return 0;
+ }
+ catch(exception& e) {
+ m->errorOut(e, "ChimeraSlayerCommand", "MPIExecute");
+ exit(1);
+ }
+}
+//**********************************************************************************************************************
+int ChimeraSlayerCommand::deconvoluteResults(SequenceParser* parser, string outputFileName, string accnosFileName, string trimFileName){
+ try {
+ map<string, string> uniqueNames = parser->getAllSeqsMap();
+ map<string, string>::iterator itUnique;
+ int total = 0;
+
+ //edit accnos file
+ ifstream in2;
+ m->openInputFile(accnosFileName, in2, "no error");
+
+ ofstream out2;
+ m->openOutputFile(accnosFileName+".temp", out2);
+
+ string name; name = "";
+ set<string> chimerasInFile;
+ set<string>::iterator itChimeras;
+
+ while (!in2.eof()) {
+ if (m->control_pressed) { in2.close(); out2.close(); m->mothurRemove(outputFileName); m->mothurRemove((accnosFileName+".temp")); return 0; }
+
+ in2 >> name; m->gobble(in2);
+
+ //find unique name
+ itUnique = uniqueNames.find(name);
+
+ if (itUnique == uniqueNames.end()) { m->mothurOut("[ERROR]: trouble parsing accnos results. Cannot find "+ name + "."); m->mothurOutEndLine(); m->control_pressed = true; }
+ else {
+ itChimeras = chimerasInFile.find((itUnique->second));
+
+ if (itChimeras == chimerasInFile.end()) {
+ out2 << itUnique->second << endl;
+ chimerasInFile.insert((itUnique->second));
+ total++;
+ }
+ }
+ }
+ in2.close();
+ out2.close();
+
+ m->mothurRemove(accnosFileName);
+ rename((accnosFileName+".temp").c_str(), accnosFileName.c_str());
+
+
+ //edit chimera file
+ ifstream in;
+ m->openInputFile(outputFileName, in);
+
+ ofstream out;
+ m->openOutputFile(outputFileName+".temp", out); out.setf(ios::fixed, ios::floatfield); out.setf(ios::showpoint);
+
+ string rest, parent1, parent2, line;
+ set<string> namesInFile; //this is so if a sequence is found to be chimera in several samples we dont write it to the results file more than once
+ set<string>::iterator itNames;
+
+ //assumptions - in file each read will always look like...
+ /*
+ F11Fcsw_92754 no
+ F11Fcsw_63104 F11Fcsw_33372 F11Fcsw_37007 0.89441 80.4469 0.2 1.03727 93.2961 52.2 no 0-241 243-369
+ */
+
+ //get header line
+ if (!in.eof()) {
+ line = m->getline(in); m->gobble(in);
+ out << line << endl;
+ }
+
+ //for the chimera file, we want to make sure if any group finds a sequence to be chimeric then all groups do,
+ //so if this is a report that did not find it to be chimeric, but it appears in the accnos file,
+ //then ignore this report and continue until we find the report that found it to be chimeric
+
+ while (!in.eof()) {
+
+ if (m->control_pressed) { in.close(); out.close(); m->mothurRemove((outputFileName+".temp")); return 0; }
+
+ in >> name; m->gobble(in);
+ in >> parent1; m->gobble(in);
+
+ if (name == "Name") { //name = "Name" because we append the header line each time we add results from the groups
+ line = m->getline(in); m->gobble(in);
+ }else {
+ if (parent1 == "no") {
+ //find unique name
+ itUnique = uniqueNames.find(name);
+
+ if (itUnique == uniqueNames.end()) { m->mothurOut("[ERROR]: trouble parsing chimera results. Cannot find "+ name + "."); m->mothurOutEndLine(); m->control_pressed = true; }
+ else {
+ //is this sequence really not chimeric??
+ itChimeras = chimerasInFile.find(itUnique->second);
+
+ if (itChimeras == chimerasInFile.end()) {
+ itNames = namesInFile.find((itUnique->second));
+
+ if (itNames == namesInFile.end()) {cout << itUnique->second << endl; out << itUnique->second << '\t' << "no" << endl; namesInFile.insert(itUnique->second); }
+ }
+ }
+ }else { //read the rest of the line
+ double DivQLAQRB,PerIDQLAQRB,BootStrapA,DivQLBQRA,PerIDQLBQRA,BootStrapB;
+ string flag, range1, range2;
+ bool print = false;
+ in >> parent2 >> DivQLAQRB >> PerIDQLAQRB >> BootStrapA >> DivQLBQRA >> PerIDQLBQRA >> BootStrapB >> flag >> range1 >> range2; m->gobble(in);
+
+ //find unique name
+ itUnique = uniqueNames.find(name);
+
+ if (itUnique == uniqueNames.end()) { m->mothurOut("[ERROR]: trouble parsing chimera results. Cannot find "+ name + "."); m->mothurOutEndLine(); m->control_pressed = true; }
+ else {
+ name = itUnique->second;
+ //is this name already in the file
+ itNames = namesInFile.find((name));
+
+ if (itNames == namesInFile.end()) { //no not in file
+ if (flag == "no") { //are you really a no??
+ //is this sequence really not chimeric??
+ itChimeras = chimerasInFile.find(name);
+
+ //then you really are a no so print, otherwise skip
+ if (itChimeras == chimerasInFile.end()) { print = true; }
+ }else{ print = true; }
+ }
+ }
+
+ if (print) {
+ out << name << '\t';
+ cout << name<< endl;
+ namesInFile.insert(name);
+
+ //output parent1's name
+ itUnique = uniqueNames.find(parent1);
+ if (itUnique == uniqueNames.end()) { m->mothurOut("[ERROR]: trouble parsing chimera results. Cannot find parentA "+ parent1 + "."); m->mothurOutEndLine(); m->control_pressed = true; }
+ else { out << itUnique->second << '\t'; }
+
+ //output parent2's name
+ itUnique = uniqueNames.find(parent2);
+ if (itUnique == uniqueNames.end()) { m->mothurOut("[ERROR]: trouble parsing chimera results. Cannot find parentA "+ parent2 + "."); m->mothurOutEndLine(); m->control_pressed = true; }
+ else { out << itUnique->second << '\t'; }
+
+ out << DivQLAQRB << '\t' << PerIDQLAQRB << '\t' << BootStrapA << '\t' << DivQLBQRA << '\t' << PerIDQLBQRA << '\t' << BootStrapB << '\t' << flag << '\t' << range1 << '\t' << range2 << endl;
+ }
+ }
+ }
+ }
+ in.close();
+ out.close();
+
+ m->mothurRemove(outputFileName);
+ rename((outputFileName+".temp").c_str(), outputFileName.c_str());
+
+ //edit fasta file
+ if (trim) {
+ ifstream in3;
+ m->openInputFile(trimFileName, in3);
+
+ ofstream out3;
+ m->openOutputFile(trimFileName+".temp", out3);
+
+ namesInFile.clear();
+
+ while (!in3.eof()) {
+ if (m->control_pressed) { in3.close(); out3.close(); m->mothurRemove(outputFileName); m->mothurRemove(accnosFileName); m->mothurRemove((trimFileName+".temp")); return 0; }
+
+ Sequence seq(in3); m->gobble(in3);
+
+ if (seq.getName() != "") {
+ //find unique name
+ itUnique = uniqueNames.find(seq.getName());
+
+ if (itUnique == uniqueNames.end()) { m->mothurOut("[ERROR]: trouble parsing accnos results. Cannot find "+ seq.getName() + "."); m->mothurOutEndLine(); m->control_pressed = true; }
+ else {
+ itNames = namesInFile.find((itUnique->second));
+
+ if (itNames == namesInFile.end()) {
+ seq.printSequence(out3);
+ }
+ }
+ }
+ }
+ in3.close();
+ out3.close();
+
+ m->mothurRemove(trimFileName);
+ rename((trimFileName+".temp").c_str(), trimFileName.c_str());
+ }
+
+ return total;
+ }
+ catch(exception& e) {
+ m->errorOut(e, "ChimeraSlayerCommand", "deconvoluteResults");
+ exit(1);
+ }
+}
+//**********************************************************************************************************************
+int ChimeraSlayerCommand::setupChimera(string inputFile, map<string, int>& priority){
+ try {
+ if (templatefile != "self") { //you want to run slayer with a reference template
+ chimera = new ChimeraSlayer(inputFile, templatefile, trim, search, ksize, match, mismatch, window, divR, minSimilarity, minCoverage, minBS, minSNP, parents, iters, increment, numwanted, realign, blastlocation, rand());
+ }else {
+ chimera = new ChimeraSlayer(inputFile, templatefile, trim, priority, search, ksize, match, mismatch, window, divR, minSimilarity, minCoverage, minBS, minSNP, parents, iters, increment, numwanted, realign, blastlocation, rand());
+ }
+
+ if (m->control_pressed) { delete chimera; return 0; }
+
+ if (chimera->getUnaligned()) { delete chimera; m->mothurOut("Your template sequences are different lengths, please correct."); m->mothurOutEndLine(); m->control_pressed = true; return 0; }
+
+ return (chimera->getLength());
+ }
+ catch(exception& e) {
+ m->errorOut(e, "ChimeraSlayerCommand", "setupChimera");
+ exit(1);
+ }
+}
+//**********************************************************************************************************************
+
+string ChimeraSlayerCommand::getNamesFile(string& inputFile){
+ try {
+ string nameFile = "";
+
+ m->mothurOutEndLine(); m->mothurOut("No namesfile given, running unique.seqs command to generate one."); m->mothurOutEndLine(); m->mothurOutEndLine();
+
+ //use unique.seqs to create new name and fastafile
+ string inputString = "fasta=" + inputFile;
+ m->mothurOut("/******************************************/"); m->mothurOutEndLine();
+ m->mothurOut("Running command: unique.seqs(" + inputString + ")"); m->mothurOutEndLine();
+
+ Command* uniqueCommand = new DeconvoluteCommand(inputString);
+ uniqueCommand->execute();
+
+ map<string, vector<string> > filenames = uniqueCommand->getOutputFiles();
+
+ delete uniqueCommand;
+
+ m->mothurOut("/******************************************/"); m->mothurOutEndLine();
+
+ nameFile = filenames["name"][0];
+ inputFile = filenames["fasta"][0];
+
+ return nameFile;
+ }
+ catch(exception& e) {
+ m->errorOut(e, "ChimeraSlayerCommand", "getNamesFile");
+ exit(1);
+ }
+}
+//**********************************************************************************************************************
int ChimeraSlayerCommand::driver(linePair* filePos, string outputFName, string filename, string accnos, string fasta){
try {
m->openInputFile(filename, inFASTA);
inFASTA.seekg(filePos->start);
+
+ if (filePos->start == 0) { chimera->printHeader(out); }
bool done = false;
int count = 0;
string candidateAligned = candidateSeq->getAligned();
if (candidateSeq->getName() != "") { //incase there is a commented sequence at the end of a file
-
if (candidateSeq->getAligned().length() != templateSeqsLength) {
m->mothurOut(candidateSeq->getName() + " is not the same length as the template sequences. Skipping."); m->mothurOutEndLine();
}else{
}
#if defined (__APPLE__) || (__MACH__) || (linux) || (__linux)
- unsigned long int pos = inFASTA.tellg();
+ unsigned long long pos = inFASTA.tellg();
if ((pos == -1) || (pos >= filePos->end)) { break; }
#else
if (inFASTA.eof()) { break; }
}
//**********************************************************************************************************************
#ifdef USE_MPI
-int ChimeraSlayerCommand::driverMPI(int start, int num, MPI_File& inMPI, MPI_File& outMPI, MPI_File& outAccMPI, MPI_File& outFastaMPI, vector<unsigned long int>& MPIPos){
+int ChimeraSlayerCommand::driverMPI(int start, int num, MPI_File& inMPI, MPI_File& outMPI, MPI_File& outAccMPI, MPI_File& outFastaMPI, vector<unsigned long long>& MPIPos){
try {
MPI_Status status;
int pid;
int ChimeraSlayerCommand::createProcesses(string outputFileName, string filename, string accnos, string fasta) {
try {
-#if defined (__APPLE__) || (__MACH__) || (linux) || (__linux)
int process = 0;
int num = 0;
+ int numNoParents = 0;
+ processIDS.clear();
+#if defined (__APPLE__) || (__MACH__) || (linux) || (__linux)
//loop through and create all the processes you want
while (process != processors) {
int pid = fork();
ofstream out;
string tempFile = outputFileName + toString(getpid()) + ".num.temp";
m->openOutputFile(tempFile, out);
- out << num << endl;
+ out << num << '\t' << chimera->getNumNoParents() << endl;
out.close();
-
exit(0);
}else {
m->mothurOut("[ERROR]: unable to spawn the necessary processes."); m->mothurOutEndLine();
ifstream in;
string tempFile = outputFileName + toString(processIDS[i]) + ".num.temp";
m->openInputFile(tempFile, in);
- if (!in.eof()) { int tempNum = 0; in >> tempNum; num += tempNum; }
- in.close(); remove(tempFile.c_str());
+ if (!in.eof()) { int tempNum = 0; int tempNumParents = 0; in >> tempNum >> tempNumParents; num += tempNum; numNoParents += tempNumParents; }
+ in.close(); m->mothurRemove(tempFile);
+ }
+#else
+
+ //////////////////////////////////////////////////////////////////////////////////////////////////////
+ //Windows version shared memory, so be careful when passing variables through the slayerData struct.
+ //Above fork() will clone, so memory is separate, but that's not the case with windows,
+ //////////////////////////////////////////////////////////////////////////////////////////////////////
+
+ vector<slayerData*> pDataArray;
+ DWORD dwThreadIdArray[processors];
+ HANDLE hThreadArray[processors];
+
+ //Create processor worker threads.
+ for( int i=0; i<processors; i++ ){
+ string extension = toString(i) + ".temp";
+ slayerData* tempslayer = new slayerData((outputFileName + extension), (fasta + extension), (accnos + extension), filename, templatefile, search, blastlocation, trimera, trim, realign, m, lines[i]->start, lines[i]->end, ksize, match, mismatch, window, minSimilarity, minCoverage, minBS, minSNP, parents, iters, increment, numwanted, divR, priority, i);
+ pDataArray.push_back(tempslayer);
+ processIDS.push_back(i);
+
+ //MySeqSumThreadFunction is in header. It must be global or static to work with the threads.
+ //default security attributes, thread function name, argument to thread function, use default creation flags, returns the thread identifier
+ hThreadArray[i] = CreateThread(NULL, 0, MySlayerThreadFunction, pDataArray[i], 0, &dwThreadIdArray[i]);
+ }
+
+ //Wait until all threads have terminated.
+ WaitForMultipleObjects(processors, hThreadArray, TRUE, INFINITE);
+
+ //Close all thread handles and free memory allocations.
+ for(int i=0; i < pDataArray.size(); i++){
+ num += pDataArray[i]->count;
+ numNoParents += pDataArray[i]->numNoParents;
+ CloseHandle(hThreadArray[i]);
+ delete pDataArray[i];
+ }
+#endif
+ if (num == numNoParents) { m->mothurOut("[WARNING]: megablast returned 0 potential parents for all your sequences. This could be due to formatdb.exe not being setup properly, please check formatdb.log for errors."); m->mothurOutEndLine(); }
+
+ rename((outputFileName + toString(processIDS[0]) + ".temp").c_str(), outputFileName.c_str());
+ rename((accnos + toString(processIDS[0]) + ".temp").c_str(), accnos.c_str());
+ if (trim) { rename((fasta + toString(processIDS[0]) + ".temp").c_str(), fasta.c_str()); }
+
+ //append output files
+ for(int i=1;i<processIDS.size();i++){
+ m->appendFiles((outputFileName + toString(processIDS[i]) + ".temp"), outputFileName);
+ m->mothurRemove((outputFileName + toString(processIDS[i]) + ".temp"));
+
+ m->appendFiles((accnos + toString(processIDS[i]) + ".temp"), accnos);
+ m->mothurRemove((accnos + toString(processIDS[i]) + ".temp"));
+
+ if (trim) {
+ m->appendFiles((fasta + toString(processIDS[i]) + ".temp"), fasta);
+ m->mothurRemove((fasta + toString(processIDS[i]) + ".temp"));
+ }
}
+
return num;
-#endif
}
catch(exception& e) {
m->errorOut(e, "ChimeraSlayerCommand", "createProcesses");
exit(1);
}
}
-
+/**************************************************************************************************/
+map<string, int> ChimeraSlayerCommand::sortFastaFile(vector<Sequence>& thisseqs, map<string, string>& nameMap, string newFile) {
+ try {
+ map<string, int> nameAbund;
+ vector<seqPriorityNode> nameVector;
+
+ //read through fastafile and store info
+ map<string, string> seqs;
+
+ for (int i = 0; i < thisseqs.size(); i++) {
+
+ if (m->control_pressed) { return nameAbund; }
+
+ map<string, string>::iterator itNameMap = nameMap.find(thisseqs[i].getName());
+
+ if (itNameMap == nameMap.end()){
+ m->control_pressed = true;
+ m->mothurOut("[ERROR]: " + thisseqs[i].getName() + " is in your fastafile, but is not in your namesfile, please correct."); m->mothurOutEndLine();
+ }else {
+ int num = m->getNumNames(itNameMap->second);
+
+ seqPriorityNode temp(num, thisseqs[i].getAligned(), thisseqs[i].getName());
+ nameVector.push_back(temp);
+ }
+ }
+
+ //sort by num represented
+ sort(nameVector.begin(), nameVector.end(), compareSeqPriorityNodes);
+
+ if (m->control_pressed) { return nameAbund; }
+
+ if (thisseqs.size() != nameVector.size()) { m->mothurOut( "The number of sequences in your fastafile does not match the number of sequences in your namefile, aborting."); m->mothurOutEndLine(); m->control_pressed = true; return nameAbund; }
+
+ ofstream out;
+ m->openOutputFile(newFile, out);
+
+ //print new file in order of
+ for (int i = 0; i < nameVector.size(); i++) {
+ out << ">" << nameVector[i].name << endl << nameVector[i].seq << endl;
+ nameAbund[nameVector[i].name] = nameVector[i].numIdentical;
+ }
+ out.close();
+
+ return nameAbund;
+
+ }
+ catch(exception& e) {
+ m->errorOut(e, "ChimeraSlayerCommand", "sortFastaFile");
+ exit(1);
+ }
+}
/**************************************************************************************************/