2 * chimeraslayercommand.cpp
5 * Created by westcott on 3/31/10.
6 * Copyright 2010 Schloss Lab. All rights reserved.
10 #include "chimeraslayercommand.h"
11 #include "chimeraslayer.h"
12 #include "deconvolutecommand.h"
14 //**********************************************************************************************************************
15 vector<string> ChimeraSlayerCommand::setParameters(){
17 CommandParameter ptemplate("reference", "InputTypes", "", "", "none", "none", "none",false,true); parameters.push_back(ptemplate);
18 CommandParameter pfasta("fasta", "InputTypes", "", "", "none", "none", "none",false,true); parameters.push_back(pfasta);
19 CommandParameter pname("name", "InputTypes", "", "", "none", "none", "none",false,false); parameters.push_back(pname);
20 CommandParameter pwindow("window", "Number", "", "50", "", "", "",false,false); parameters.push_back(pwindow);
21 CommandParameter pksize("ksize", "Number", "", "7", "", "", "",false,false); parameters.push_back(pksize);
22 CommandParameter pmatch("match", "Number", "", "5.0", "", "", "",false,false); parameters.push_back(pmatch);
23 CommandParameter pmismatch("mismatch", "Number", "", "-4.0", "", "", "",false,false); parameters.push_back(pmismatch);
24 CommandParameter pminsim("minsim", "Number", "", "90", "", "", "",false,false); parameters.push_back(pminsim);
25 CommandParameter pmincov("mincov", "Number", "", "70", "", "", "",false,false); parameters.push_back(pmincov);
26 CommandParameter pminsnp("minsnp", "Number", "", "10", "", "", "",false,false); parameters.push_back(pminsnp);
27 CommandParameter pminbs("minbs", "Number", "", "90", "", "", "",false,false); parameters.push_back(pminbs);
28 CommandParameter psearch("search", "Multiple", "kmer-blast", "blast", "", "", "",false,false); parameters.push_back(psearch);
29 CommandParameter pprocessors("processors", "Number", "", "1", "", "", "",false,false); parameters.push_back(pprocessors);
30 CommandParameter prealign("realign", "Boolean", "", "T", "", "", "",false,false); parameters.push_back(prealign);
31 CommandParameter ptrim("trim", "Boolean", "", "F", "", "", "",false,false); parameters.push_back(ptrim);
32 CommandParameter psplit("split", "Boolean", "", "F", "", "", "",false,false); parameters.push_back(psplit);
33 CommandParameter pnumwanted("numwanted", "Number", "", "15", "", "", "",false,false); parameters.push_back(pnumwanted);
34 CommandParameter piters("iters", "Number", "", "1000", "", "", "",false,false); parameters.push_back(piters);
35 CommandParameter pdivergence("divergence", "Number", "", "1.007", "", "", "",false,false); parameters.push_back(pdivergence);
36 CommandParameter pparents("parents", "Number", "", "3", "", "", "",false,false); parameters.push_back(pparents);
37 CommandParameter pincrement("increment", "Number", "", "5", "", "", "",false,false); parameters.push_back(pincrement);
38 CommandParameter pinputdir("inputdir", "String", "", "", "", "", "",false,false); parameters.push_back(pinputdir);
39 CommandParameter poutputdir("outputdir", "String", "", "", "", "", "",false,false); parameters.push_back(poutputdir);
41 vector<string> myArray;
42 for (int i = 0; i < parameters.size(); i++) { myArray.push_back(parameters[i].name); }
46 m->errorOut(e, "ChimeraSlayerCommand", "setParameters");
50 //**********************************************************************************************************************
51 string ChimeraSlayerCommand::getHelpString(){
53 string helpString = "";
54 helpString += "The chimera.slayer command reads a fastafile and referencefile and outputs potentially chimeric sequences.\n";
55 helpString += "This command was modeled after the chimeraSlayer written by the Broad Institute.\n";
56 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";
57 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";
58 helpString += "The name parameter allows you to provide a name file, if you are using template=self. \n";
59 helpString += "You may enter multiple fasta files by separating their names with dashes. ie. fasta=abrecovery.fasta-amazon.fasta \n";
60 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";
61 helpString += "The processors parameter allows you to specify how many processors you would like to use. The default is 1. \n";
63 helpString += "When using MPI, the processors parameter is set to the number of MPI processes running. \n";
65 helpString += "The trim parameter allows you to output a new fasta file containing your sequences with the chimeric ones trimmed to include only their longest piece, default=F. \n";
66 helpString += "The split parameter allows you to check both pieces of non-chimeric sequence for chimeras, thus looking for trimeras and quadmeras. default=F. \n";
67 helpString += "The window parameter allows you to specify the window size for searching for chimeras, default=50. \n";
68 helpString += "The increment parameter allows you to specify how far you move each window while finding chimeric sequences, default=5.\n";
69 helpString += "The numwanted parameter allows you to specify how many sequences you would each query sequence compared with, default=15.\n";
70 helpString += "The ksize parameter allows you to input kmersize, default is 7, used if search is kmer. \n";
71 helpString += "The match parameter allows you to reward matched bases in blast search, default is 5. \n";
72 helpString += "The parents parameter allows you to select the number of potential parents to investigate from the numwanted best matches after rating them, default is 3. \n";
73 helpString += "The mismatch parameter allows you to penalize mismatched bases in blast search, default is -4. \n";
74 helpString += "The divergence parameter allows you to set a cutoff for chimera determination, default is 1.007. \n";
75 helpString += "The iters parameter allows you to specify the number of bootstrap iters to do with the chimeraslayer method, default=1000.\n";
76 helpString += "The minsim parameter allows you to specify a minimum similarity with the parent fragments, default=90. \n";
77 helpString += "The mincov parameter allows you to specify minimum coverage by closest matches found in template. Default is 70, meaning 70%. \n";
78 helpString += "The minbs parameter allows you to specify minimum bootstrap support for calling a sequence chimeric. Default is 90, meaning 90%. \n";
79 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";
80 helpString += "The search parameter allows you to specify search method for finding the closest parent. Choices are blast, and kmer, default blast. \n";
81 helpString += "The realign parameter allows you to realign the query to the potential parents. Choices are true or false, default true. \n";
82 helpString += "The chimera.slayer command should be in the following format: \n";
83 helpString += "chimera.slayer(fasta=yourFastaFile, reference=yourTemplate, search=yourSearch) \n";
84 helpString += "Example: chimera.slayer(fasta=AD.align, reference=core_set_aligned.imputed.fasta, search=kmer) \n";
85 helpString += "Note: No spaces between parameter labels (i.e. fasta), '=' and parameters (i.e.yourFastaFile).\n";
89 m->errorOut(e, "ChimeraSlayerCommand", "getHelpString");
93 //**********************************************************************************************************************
94 ChimeraSlayerCommand::ChimeraSlayerCommand(){
96 abort = true; calledHelp = true;
98 vector<string> tempOutNames;
99 outputTypes["chimera"] = tempOutNames;
100 outputTypes["accnos"] = tempOutNames;
101 outputTypes["fasta"] = tempOutNames;
103 catch(exception& e) {
104 m->errorOut(e, "ChimeraSlayerCommand", "ChimeraSlayerCommand");
108 //***************************************************************************************************************
109 ChimeraSlayerCommand::ChimeraSlayerCommand(string option) {
111 abort = false; calledHelp = false;
113 //allow user to run help
114 if(option == "help") { help(); abort = true; calledHelp = true; }
115 else if(option == "citation") { citation(); abort = true; calledHelp = true;}
118 vector<string> myArray = setParameters();
120 OptionParser parser(option);
121 map<string,string> parameters = parser.getParameters();
123 ValidParameters validParameter("chimera.slayer");
124 map<string,string>::iterator it;
126 //check to make sure all parameters are valid for command
127 for (it = parameters.begin(); it != parameters.end(); it++) {
128 if (validParameter.isValidParameter(it->first, myArray, it->second) != true) { abort = true; }
131 vector<string> tempOutNames;
132 outputTypes["chimera"] = tempOutNames;
133 outputTypes["accnos"] = tempOutNames;
134 outputTypes["fasta"] = tempOutNames;
136 //if the user changes the input directory command factory will send this info to us in the output parameter
137 string inputDir = validParameter.validFile(parameters, "inputdir", false);
138 if (inputDir == "not found"){ inputDir = ""; }
140 //check for required parameters
141 fastafile = validParameter.validFile(parameters, "fasta", false);
142 if (fastafile == "not found") {
143 //if there is a current fasta file, use it
144 string filename = m->getFastaFile();
145 if (filename != "") { fastaFileNames.push_back(filename); m->mothurOut("Using " + filename + " as input file for the fasta parameter."); m->mothurOutEndLine(); }
146 else { m->mothurOut("You have no current fastafile and the fasta parameter is required."); m->mothurOutEndLine(); abort = true; }
148 m->splitAtDash(fastafile, fastaFileNames);
150 //go through files and make sure they are good, if not, then disregard them
151 for (int i = 0; i < fastaFileNames.size(); i++) {
154 if (fastaFileNames[i] == "current") {
155 fastaFileNames[i] = m->getFastaFile();
156 if (fastaFileNames[i] != "") { m->mothurOut("Using " + fastaFileNames[i] + " as input file for the fasta parameter where you had given current."); m->mothurOutEndLine(); }
158 m->mothurOut("You have no current fastafile, ignoring current."); m->mothurOutEndLine(); ignore=true;
159 //erase from file list
160 fastaFileNames.erase(fastaFileNames.begin()+i);
167 if (inputDir != "") {
168 string path = m->hasPath(fastaFileNames[i]);
169 //if the user has not given a path then, add inputdir. else leave path alone.
170 if (path == "") { fastaFileNames[i] = inputDir + fastaFileNames[i]; }
176 ableToOpen = m->openInputFile(fastaFileNames[i], in, "noerror");
178 //if you can't open it, try default location
179 if (ableToOpen == 1) {
180 if (m->getDefaultPath() != "") { //default path is set
181 string tryPath = m->getDefaultPath() + m->getSimpleName(fastaFileNames[i]);
182 m->mothurOut("Unable to open " + fastaFileNames[i] + ". Trying default " + tryPath); m->mothurOutEndLine();
184 ableToOpen = m->openInputFile(tryPath, in2, "noerror");
186 fastaFileNames[i] = tryPath;
190 if (ableToOpen == 1) {
191 if (m->getOutputDir() != "") { //default path is set
192 string tryPath = m->getOutputDir() + m->getSimpleName(fastaFileNames[i]);
193 m->mothurOut("Unable to open " + fastaFileNames[i] + ". Trying output directory " + tryPath); m->mothurOutEndLine();
195 ableToOpen = m->openInputFile(tryPath, in2, "noerror");
197 fastaFileNames[i] = tryPath;
203 if (ableToOpen == 1) {
204 m->mothurOut("Unable to open " + fastaFileNames[i] + ". It will be disregarded."); m->mothurOutEndLine();
205 //erase from file list
206 fastaFileNames.erase(fastaFileNames.begin()+i);
212 //make sure there is at least one valid file left
213 if (fastaFileNames.size() == 0) { m->mothurOut("[ERROR]: no valid files."); m->mothurOutEndLine(); abort = true; }
217 //check for required parameters
219 namefile = validParameter.validFile(parameters, "name", false);
220 if (namefile == "not found") { namefile = ""; hasName = false; }
222 m->splitAtDash(namefile, nameFileNames);
224 //go through files and make sure they are good, if not, then disregard them
225 for (int i = 0; i < nameFileNames.size(); i++) {
228 if (nameFileNames[i] == "current") {
229 nameFileNames[i] = m->getNameFile();
230 if (nameFileNames[i] != "") { m->mothurOut("Using " + nameFileNames[i] + " as input file for the name parameter where you had given current."); m->mothurOutEndLine(); }
232 m->mothurOut("You have no current namefile, ignoring current."); m->mothurOutEndLine(); ignore=true;
233 //erase from file list
234 nameFileNames.erase(nameFileNames.begin()+i);
241 if (inputDir != "") {
242 string path = m->hasPath(nameFileNames[i]);
243 //if the user has not given a path then, add inputdir. else leave path alone.
244 if (path == "") { nameFileNames[i] = inputDir + nameFileNames[i]; }
250 ableToOpen = m->openInputFile(nameFileNames[i], in, "noerror");
252 //if you can't open it, try default location
253 if (ableToOpen == 1) {
254 if (m->getDefaultPath() != "") { //default path is set
255 string tryPath = m->getDefaultPath() + m->getSimpleName(nameFileNames[i]);
256 m->mothurOut("Unable to open " + nameFileNames[i] + ". Trying default " + tryPath); m->mothurOutEndLine();
258 ableToOpen = m->openInputFile(tryPath, in2, "noerror");
260 nameFileNames[i] = tryPath;
264 if (ableToOpen == 1) {
265 if (m->getOutputDir() != "") { //default path is set
266 string tryPath = m->getOutputDir() + m->getSimpleName(nameFileNames[i]);
267 m->mothurOut("Unable to open " + nameFileNames[i] + ". Trying output directory " + tryPath); m->mothurOutEndLine();
269 ableToOpen = m->openInputFile(tryPath, in2, "noerror");
271 nameFileNames[i] = tryPath;
277 if (ableToOpen == 1) {
278 m->mothurOut("Unable to open " + nameFileNames[i] + ". It will be disregarded."); m->mothurOutEndLine();
279 //erase from file list
280 nameFileNames.erase(nameFileNames.begin()+i);
286 //make sure there is at least one valid file left
287 if (nameFileNames.size() == 0) { m->mothurOut("[ERROR]: no valid name files."); m->mothurOutEndLine(); abort = true; }
290 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; }
292 //if the user changes the output directory command factory will send this info to us in the output parameter
293 outputDir = validParameter.validFile(parameters, "outputdir", false); if (outputDir == "not found"){ outputDir = ""; }
297 it = parameters.find("reference");
298 //user has given a template file
299 if(it != parameters.end()){
300 if (it->second == "self") { templatefile = "self"; }
302 path = m->hasPath(it->second);
303 //if the user has not given a path then, add inputdir. else leave path alone.
304 if (path == "") { parameters["reference"] = inputDir + it->second; }
306 templatefile = validParameter.validFile(parameters, "reference", true);
307 if (templatefile == "not open") { abort = true; }
308 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; }
310 }else if (hasName) { templatefile = "self"; }
311 else { templatefile = ""; m->mothurOut("reference is a required parameter for the chimera.slayer command."); m->mothurOutEndLine(); abort = true; }
313 string temp = validParameter.validFile(parameters, "processors", false); if (temp == "not found"){ temp = m->getProcessors(); }
314 m->setProcessors(temp);
315 convert(temp, processors);
317 temp = validParameter.validFile(parameters, "ksize", false); if (temp == "not found") { temp = "7"; }
318 convert(temp, ksize);
320 temp = validParameter.validFile(parameters, "window", false); if (temp == "not found") { temp = "50"; }
321 convert(temp, window);
323 temp = validParameter.validFile(parameters, "match", false); if (temp == "not found") { temp = "5"; }
324 convert(temp, match);
326 temp = validParameter.validFile(parameters, "mismatch", false); if (temp == "not found") { temp = "-4"; }
327 convert(temp, mismatch);
329 temp = validParameter.validFile(parameters, "divergence", false); if (temp == "not found") { temp = "1.007"; }
332 temp = validParameter.validFile(parameters, "minsim", false); if (temp == "not found") { temp = "90"; }
333 convert(temp, minSimilarity);
335 temp = validParameter.validFile(parameters, "mincov", false); if (temp == "not found") { temp = "70"; }
336 convert(temp, minCoverage);
338 temp = validParameter.validFile(parameters, "minbs", false); if (temp == "not found") { temp = "90"; }
339 convert(temp, minBS);
341 temp = validParameter.validFile(parameters, "minsnp", false); if (temp == "not found") { temp = "10"; }
342 convert(temp, minSNP);
344 temp = validParameter.validFile(parameters, "parents", false); if (temp == "not found") { temp = "3"; }
345 convert(temp, parents);
347 temp = validParameter.validFile(parameters, "realign", false); if (temp == "not found") { temp = "t"; }
348 realign = m->isTrue(temp);
350 temp = validParameter.validFile(parameters, "trim", false); if (temp == "not found") { temp = "f"; }
351 trim = m->isTrue(temp);
353 temp = validParameter.validFile(parameters, "split", false); if (temp == "not found") { temp = "f"; }
354 trimera = m->isTrue(temp);
356 search = validParameter.validFile(parameters, "search", false); if (search == "not found") { search = "blast"; }
358 temp = validParameter.validFile(parameters, "iters", false); if (temp == "not found") { temp = "1000"; }
359 convert(temp, iters);
361 temp = validParameter.validFile(parameters, "increment", false); if (temp == "not found") { temp = "5"; }
362 convert(temp, increment);
364 temp = validParameter.validFile(parameters, "numwanted", false); if (temp == "not found") { temp = "15"; }
365 convert(temp, numwanted);
367 if ((search != "blast") && (search != "kmer")) { m->mothurOut(search + " is not a valid search."); m->mothurOutEndLine(); abort = true; }
370 catch(exception& e) {
371 m->errorOut(e, "ChimeraSlayerCommand", "ChimeraSlayerCommand");
375 //***************************************************************************************************************
377 int ChimeraSlayerCommand::execute(){
379 if (abort == true) { if (calledHelp) { return 0; } return 2; }
381 for (int s = 0; s < fastaFileNames.size(); s++) {
383 m->mothurOut("Checking sequences from " + fastaFileNames[s] + " ..." ); m->mothurOutEndLine();
385 int start = time(NULL);
387 if (templatefile != "self") { //you want to run slayer with a refernce template
388 chimera = new ChimeraSlayer(fastaFileNames[s], templatefile, trim, search, ksize, match, mismatch, window, divR, minSimilarity, minCoverage, minBS, minSNP, parents, iters, increment, numwanted, realign);
390 if (processors != 1) { m->mothurOut("When using template=self, mothur can only use 1 processor, continuing."); m->mothurOutEndLine(); processors = 1; }
391 string nameFile = "";
392 if (nameFileNames.size() != 0) { //you provided a namefile and we don't need to create one
393 nameFile = nameFileNames[s];
395 m->mothurOutEndLine(); m->mothurOut("No namesfile given, running unique.seqs command to generate one."); m->mothurOutEndLine(); m->mothurOutEndLine();
397 //use unique.seqs to create new name and fastafile
398 string inputString = "fasta=" + fastaFileNames[s];
399 m->mothurOut("/******************************************/"); m->mothurOutEndLine();
400 m->mothurOut("Running command: unique.seqs(" + inputString + ")"); m->mothurOutEndLine();
402 Command* uniqueCommand = new DeconvoluteCommand(inputString);
403 uniqueCommand->execute();
405 map<string, vector<string> > filenames = uniqueCommand->getOutputFiles();
407 delete uniqueCommand;
409 m->mothurOut("/******************************************/"); m->mothurOutEndLine();
411 nameFile = filenames["name"][0];
412 fastaFileNames[s] = filenames["fasta"][0];
415 //sort fastafile by abundance, returns new sorted fastafile name
416 m->mothurOut("Sorting fastafile according to abundance..."); cout.flush();
417 map<string, int> priority = sortFastaFile(fastaFileNames[s], nameFile);
418 m->mothurOut("Done."); m->mothurOutEndLine();
420 if (m->control_pressed) { for (int j = 0; j < outputNames.size(); j++) { remove(outputNames[j].c_str()); } return 0; }
422 chimera = new ChimeraSlayer(fastaFileNames[s], templatefile, trim, priority, search, ksize, match, mismatch, window, divR, minSimilarity, minCoverage, minBS, minSNP, parents, iters, increment, numwanted, realign);
425 if (outputDir == "") { outputDir = m->hasPath(fastaFileNames[s]); }//if user entered a file with a path then preserve it
426 string outputFileName = outputDir + m->getRootName(m->getSimpleName(fastaFileNames[s])) + "slayer.chimera";
427 string accnosFileName = outputDir + m->getRootName(m->getSimpleName(fastaFileNames[s])) + "slayer.accnos";
428 string trimFastaFileName = outputDir + m->getRootName(m->getSimpleName(fastaFileNames[s])) + "slayer.fasta";
430 if (m->control_pressed) { delete chimera; for (int j = 0; j < outputNames.size(); j++) { remove(outputNames[j].c_str()); } return 0; }
432 if (chimera->getUnaligned()) {
433 m->mothurOut("Your template sequences are different lengths, please correct."); m->mothurOutEndLine();
437 templateSeqsLength = chimera->getLength();
440 int pid, numSeqsPerProcessor;
442 vector<unsigned long int> MPIPos;
445 MPI_Comm_rank(MPI_COMM_WORLD, &pid); //find out who we are
446 MPI_Comm_size(MPI_COMM_WORLD, &processors);
450 MPI_File outMPIAccnos;
451 MPI_File outMPIFasta;
453 int outMode=MPI_MODE_CREATE|MPI_MODE_WRONLY;
454 int inMode=MPI_MODE_RDONLY;
456 char outFilename[1024];
457 strcpy(outFilename, outputFileName.c_str());
459 char outAccnosFilename[1024];
460 strcpy(outAccnosFilename, accnosFileName.c_str());
462 char outFastaFilename[1024];
463 strcpy(outFastaFilename, trimFastaFileName.c_str());
465 char inFileName[1024];
466 strcpy(inFileName, fastaFileNames[s].c_str());
468 MPI_File_open(MPI_COMM_WORLD, inFileName, inMode, MPI_INFO_NULL, &inMPI); //comm, filename, mode, info, filepointer
469 MPI_File_open(MPI_COMM_WORLD, outFilename, outMode, MPI_INFO_NULL, &outMPI);
470 MPI_File_open(MPI_COMM_WORLD, outAccnosFilename, outMode, MPI_INFO_NULL, &outMPIAccnos);
471 if (trim) { MPI_File_open(MPI_COMM_WORLD, outFastaFilename, outMode, MPI_INFO_NULL, &outMPIFasta); }
473 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; }
475 if (pid == 0) { //you are the root process
476 m->mothurOutEndLine();
477 m->mothurOut("Only reporting sequence supported by " + toString(minBS) + "% of bootstrapped results.");
478 m->mothurOutEndLine();
480 string outTemp = "Name\tLeftParent\tRightParent\tDivQLAQRB\tPerIDQLAQRB\tBootStrapA\tDivQLBQRA\tPerIDQLBQRA\tBootStrapB\tFlag\tLeftWindow\tRightWindow\n";
483 int length = outTemp.length();
484 char* buf2 = new char[length];
485 memcpy(buf2, outTemp.c_str(), length);
487 MPI_File_write_shared(outMPI, buf2, length, MPI_CHAR, &status);
490 MPIPos = m->setFilePosFasta(fastaFileNames[s], numSeqs); //fills MPIPos, returns numSeqs
492 if (templatefile != "self") { //if template=self we can only use 1 processor
493 //send file positions to all processes
494 for(int i = 1; i < processors; i++) {
495 MPI_Send(&numSeqs, 1, MPI_INT, i, tag, MPI_COMM_WORLD);
496 MPI_Send(&MPIPos[0], (numSeqs+1), MPI_LONG, i, tag, MPI_COMM_WORLD);
499 //figure out how many sequences you have to align
500 numSeqsPerProcessor = numSeqs / processors;
501 int startIndex = pid * numSeqsPerProcessor;
502 if(pid == (processors - 1)){ numSeqsPerProcessor = numSeqs - pid * numSeqsPerProcessor; }
504 if (templatefile == "self") { //if template=self we can only use 1 processor
506 numSeqsPerProcessor = numSeqs;
510 driverMPI(startIndex, numSeqsPerProcessor, inMPI, outMPI, outMPIAccnos, outMPIFasta, MPIPos);
512 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; }
514 }else{ //you are a child process
515 if (templatefile != "self") { //if template=self we can only use 1 processor
516 MPI_Recv(&numSeqs, 1, MPI_INT, 0, tag, MPI_COMM_WORLD, &status);
517 MPIPos.resize(numSeqs+1);
518 MPI_Recv(&MPIPos[0], (numSeqs+1), MPI_LONG, 0, tag, MPI_COMM_WORLD, &status);
520 //figure out how many sequences you have to align
521 numSeqsPerProcessor = numSeqs / processors;
522 int startIndex = pid * numSeqsPerProcessor;
523 if(pid == (processors - 1)){ numSeqsPerProcessor = numSeqs - pid * numSeqsPerProcessor; }
526 driverMPI(startIndex, numSeqsPerProcessor, inMPI, outMPI, outMPIAccnos, outMPIFasta, MPIPos);
528 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; }
534 MPI_File_close(&inMPI);
535 MPI_File_close(&outMPI);
536 MPI_File_close(&outMPIAccnos);
537 if (trim) { MPI_File_close(&outMPIFasta); }
538 MPI_Barrier(MPI_COMM_WORLD); //make everyone wait - just in case
542 string tempHeader = outputDir + m->getRootName(m->getSimpleName(fastaFileNames[s])) + "slayer.chimeras.tempHeader";
543 m->openOutputFile(tempHeader, outHeader);
545 chimera->printHeader(outHeader);
548 vector<unsigned long int> positions = m->divideFile(fastaFileNames[s], processors);
550 for (int i = 0; i < (positions.size()-1); i++) {
551 lines.push_back(new linePair(positions[i], positions[(i+1)]));
555 #if defined (__APPLE__) || (__MACH__) || (linux) || (__linux)
557 numSeqs = driver(lines[0], outputFileName, fastaFileNames[s], accnosFileName, trimFastaFileName);
559 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; }
562 processIDS.resize(0);
564 numSeqs = createProcesses(outputFileName, fastaFileNames[s], accnosFileName, trimFastaFileName);
566 rename((outputFileName + toString(processIDS[0]) + ".temp").c_str(), outputFileName.c_str());
567 rename((accnosFileName + toString(processIDS[0]) + ".temp").c_str(), accnosFileName.c_str());
568 if (trim) { rename((trimFastaFileName + toString(processIDS[0]) + ".temp").c_str(), trimFastaFileName.c_str()); }
570 //append output files
571 for(int i=1;i<processors;i++){
572 m->appendFiles((outputFileName + toString(processIDS[i]) + ".temp"), outputFileName);
573 remove((outputFileName + toString(processIDS[i]) + ".temp").c_str());
576 //append output files
577 for(int i=1;i<processors;i++){
578 m->appendFiles((accnosFileName + toString(processIDS[i]) + ".temp"), accnosFileName);
579 remove((accnosFileName + toString(processIDS[i]) + ".temp").c_str());
583 for(int i=1;i<processors;i++){
584 m->appendFiles((trimFastaFileName + toString(processIDS[i]) + ".temp"), trimFastaFileName);
585 remove((trimFastaFileName + toString(processIDS[i]) + ".temp").c_str());
589 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; }
593 numSeqs = driver(lines[0], outputFileName, fastaFileNames[s], accnosFileName, trimFastaFileName);
595 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; }
599 m->appendFiles(outputFileName, tempHeader);
601 remove(outputFileName.c_str());
602 rename(tempHeader.c_str(), outputFileName.c_str());
608 for (int i = 0; i < lines.size(); i++) { delete lines[i]; } lines.clear();
610 outputNames.push_back(outputFileName); outputTypes["chimera"].push_back(outputFileName);
611 outputNames.push_back(accnosFileName); outputTypes["accnos"].push_back(accnosFileName);
612 if (trim) { outputNames.push_back(trimFastaFileName); outputTypes["fasta"].push_back(trimFastaFileName); }
614 m->mothurOutEndLine(); m->mothurOut("It took " + toString(time(NULL) - start) + " secs to check " + toString(numSeqs) + " sequences."); m->mothurOutEndLine();
617 //set accnos file as new current accnosfile
619 itTypes = outputTypes.find("accnos");
620 if (itTypes != outputTypes.end()) {
621 if ((itTypes->second).size() != 0) { current = (itTypes->second)[0]; m->setAccnosFile(current); }
625 itTypes = outputTypes.find("fasta");
626 if (itTypes != outputTypes.end()) {
627 if ((itTypes->second).size() != 0) { current = (itTypes->second)[0]; m->setFastaFile(current); }
631 m->mothurOutEndLine();
632 m->mothurOut("Output File Names: "); m->mothurOutEndLine();
633 for (int i = 0; i < outputNames.size(); i++) { m->mothurOut(outputNames[i]); m->mothurOutEndLine(); }
634 m->mothurOutEndLine();
639 catch(exception& e) {
640 m->errorOut(e, "ChimeraSlayerCommand", "execute");
644 //**********************************************************************************************************************
646 int ChimeraSlayerCommand::driver(linePair* filePos, string outputFName, string filename, string accnos, string fasta){
649 m->openOutputFile(outputFName, out);
652 m->openOutputFile(accnos, out2);
655 if (trim) { m->openOutputFile(fasta, out3); }
658 m->openInputFile(filename, inFASTA);
660 inFASTA.seekg(filePos->start);
667 if (m->control_pressed) { out.close(); out2.close(); if (trim) { out3.close(); } inFASTA.close(); return 1; }
669 Sequence* candidateSeq = new Sequence(inFASTA); m->gobble(inFASTA);
670 string candidateAligned = candidateSeq->getAligned();
672 if (candidateSeq->getName() != "") { //incase there is a commented sequence at the end of a file
674 if (candidateSeq->getAligned().length() != templateSeqsLength) {
675 m->mothurOut(candidateSeq->getName() + " is not the same length as the template sequences. Skipping."); m->mothurOutEndLine();
678 chimera->getChimeras(candidateSeq);
680 if (m->control_pressed) { delete candidateSeq; return 1; }
682 //if you are not chimeric, then check each half
683 data_results wholeResults = chimera->getResults();
685 //determine if we need to split
686 bool isChimeric = false;
688 if (wholeResults.flag == "yes") {
689 string chimeraFlag = "no";
690 if( (wholeResults.results[0].bsa >= minBS && wholeResults.results[0].divr_qla_qrb >= divR)
692 (wholeResults.results[0].bsb >= minBS && wholeResults.results[0].divr_qlb_qra >= divR) ) { chimeraFlag = "yes"; }
695 if (chimeraFlag == "yes") {
696 if ((wholeResults.results[0].bsa >= minBS) || (wholeResults.results[0].bsb >= minBS)) { isChimeric = true; }
700 if ((!isChimeric) && trimera) {
702 //split sequence in half by bases
703 string leftQuery, rightQuery;
704 Sequence tempSeq(candidateSeq->getName(), candidateAligned);
705 divideInHalf(tempSeq, leftQuery, rightQuery);
707 //run chimeraSlayer on each piece
708 Sequence* left = new Sequence(candidateSeq->getName(), leftQuery);
709 Sequence* right = new Sequence(candidateSeq->getName(), rightQuery);
712 chimera->getChimeras(left);
713 data_results leftResults = chimera->getResults();
715 chimera->getChimeras(right);
716 data_results rightResults = chimera->getResults();
718 //if either piece is chimeric then report
719 Sequence* trimmed = chimera->print(out, out2, leftResults, rightResults);
720 if (trim) { trimmed->printSequence(out3); delete trimmed; }
722 delete left; delete right;
724 }else { //already chimeric
726 Sequence* trimmed = chimera->print(out, out2);
727 if (trim) { trimmed->printSequence(out3); delete trimmed; }
735 #if defined (__APPLE__) || (__MACH__) || (linux) || (__linux)
736 unsigned long int pos = inFASTA.tellg();
737 if ((pos == -1) || (pos >= filePos->end)) { break; }
739 if (inFASTA.eof()) { break; }
744 if((count) % 100 == 0){ m->mothurOut("Processing sequence: " + toString(count)); m->mothurOutEndLine(); }
747 if((count) % 100 != 0){ m->mothurOut("Processing sequence: " + toString(count)); m->mothurOutEndLine(); }
751 if (trim) { out3.close(); }
756 catch(exception& e) {
757 m->errorOut(e, "ChimeraSlayerCommand", "driver");
761 //**********************************************************************************************************************
763 int ChimeraSlayerCommand::driverMPI(int start, int num, MPI_File& inMPI, MPI_File& outMPI, MPI_File& outAccMPI, MPI_File& outFastaMPI, vector<unsigned long int>& MPIPos){
767 MPI_Comm_rank(MPI_COMM_WORLD, &pid); //find out who we are
769 for(int i=0;i<num;i++){
771 if (m->control_pressed) { return 1; }
774 int length = MPIPos[start+i+1] - MPIPos[start+i];
776 char* buf4 = new char[length];
777 MPI_File_read_at(inMPI, MPIPos[start+i], buf4, length, MPI_CHAR, &status);
779 string tempBuf = buf4;
780 if (tempBuf.length() > length) { tempBuf = tempBuf.substr(0, length); }
781 istringstream iss (tempBuf,istringstream::in);
785 Sequence* candidateSeq = new Sequence(iss); m->gobble(iss);
786 string candidateAligned = candidateSeq->getAligned();
788 if (candidateSeq->getName() != "") { //incase there is a commented sequence at the end of a file
790 if (candidateSeq->getAligned().length() != templateSeqsLength) {
791 m->mothurOut(candidateSeq->getName() + " is not the same length as the template sequences. Skipping."); m->mothurOutEndLine();
795 chimera->getChimeras(candidateSeq);
797 if (m->control_pressed) { delete candidateSeq; return 1; }
799 //if you are not chimeric, then check each half
800 data_results wholeResults = chimera->getResults();
802 //determine if we need to split
803 bool isChimeric = false;
805 if (wholeResults.flag == "yes") {
806 string chimeraFlag = "no";
807 if( (wholeResults.results[0].bsa >= minBS && wholeResults.results[0].divr_qla_qrb >= divR)
809 (wholeResults.results[0].bsb >= minBS && wholeResults.results[0].divr_qlb_qra >= divR) ) { chimeraFlag = "yes"; }
812 if (chimeraFlag == "yes") {
813 if ((wholeResults.results[0].bsa >= minBS) || (wholeResults.results[0].bsb >= minBS)) { isChimeric = true; }
817 if ((!isChimeric) && trimera) {
818 //split sequence in half by bases
819 string leftQuery, rightQuery;
820 Sequence tempSeq(candidateSeq->getName(), candidateAligned);
821 divideInHalf(tempSeq, leftQuery, rightQuery);
823 //run chimeraSlayer on each piece
824 Sequence* left = new Sequence(candidateSeq->getName(), leftQuery);
825 Sequence* right = new Sequence(candidateSeq->getName(), rightQuery);
828 chimera->getChimeras(left);
829 data_results leftResults = chimera->getResults();
831 chimera->getChimeras(right);
832 data_results rightResults = chimera->getResults();
834 //if either piece is chimeric then report
835 Sequence* trimmed = chimera->print(outMPI, outAccMPI, leftResults, rightResults);
837 string outputString = ">" + trimmed->getName() + "\n" + trimmed->getAligned() + "\n";
840 //write to accnos file
841 int length = outputString.length();
842 char* buf2 = new char[length];
843 memcpy(buf2, outputString.c_str(), length);
845 MPI_File_write_shared(outFastaMPI, buf2, length, MPI_CHAR, &status);
849 delete left; delete right;
853 Sequence* trimmed = chimera->print(outMPI, outAccMPI);
856 string outputString = ">" + trimmed->getName() + "\n" + trimmed->getAligned() + "\n";
859 //write to accnos file
860 int length = outputString.length();
861 char* buf2 = new char[length];
862 memcpy(buf2, outputString.c_str(), length);
864 MPI_File_write_shared(outFastaMPI, buf2, length, MPI_CHAR, &status);
874 if((i+1) % 100 == 0){ cout << "Processing sequence: " << (i+1) << endl; m->mothurOutJustToLog("Processing sequence: " + toString(i+1) + "\n"); }
877 if(num % 100 != 0){ cout << "Processing sequence: " << num << endl; m->mothurOutJustToLog("Processing sequence: " + toString(num) + "\n"); }
882 catch(exception& e) {
883 m->errorOut(e, "ChimeraSlayerCommand", "driverMPI");
889 /**************************************************************************************************/
891 int ChimeraSlayerCommand::createProcesses(string outputFileName, string filename, string accnos, string fasta) {
893 #if defined (__APPLE__) || (__MACH__) || (linux) || (__linux)
897 //loop through and create all the processes you want
898 while (process != processors) {
902 processIDS.push_back(pid); //create map from line number to pid so you can append files in correct order later
905 num = driver(lines[process], outputFileName + toString(getpid()) + ".temp", filename, accnos + toString(getpid()) + ".temp", fasta + toString(getpid()) + ".temp");
907 //pass numSeqs to parent
909 string tempFile = outputFileName + toString(getpid()) + ".num.temp";
910 m->openOutputFile(tempFile, out);
916 m->mothurOut("[ERROR]: unable to spawn the necessary processes."); m->mothurOutEndLine();
917 for (int i = 0; i < processIDS.size(); i++) { kill (processIDS[i], SIGINT); }
922 //force parent to wait until all the processes are done
923 for (int i=0;i<processors;i++) {
924 int temp = processIDS[i];
928 for (int i = 0; i < processIDS.size(); i++) {
930 string tempFile = outputFileName + toString(processIDS[i]) + ".num.temp";
931 m->openInputFile(tempFile, in);
932 if (!in.eof()) { int tempNum = 0; in >> tempNum; num += tempNum; }
933 in.close(); remove(tempFile.c_str());
939 catch(exception& e) {
940 m->errorOut(e, "ChimeraSlayerCommand", "createProcesses");
945 /**************************************************************************************************/
947 int ChimeraSlayerCommand::divideInHalf(Sequence querySeq, string& leftQuery, string& rightQuery) {
950 string queryUnAligned = querySeq.getUnaligned();
951 int numBases = int(queryUnAligned.length() * 0.5);
953 string queryAligned = querySeq.getAligned();
954 leftQuery = querySeq.getAligned();
955 rightQuery = querySeq.getAligned();
959 for (int i = 0; i < queryAligned.length(); i++) {
961 if (isalpha(queryAligned[i])) {
966 if (baseCount >= numBases) { leftSpot = i; break; } //first half
969 //blank out right side
970 for (int i = leftSpot; i < leftQuery.length(); i++) { leftQuery[i] = '.'; }
972 //blank out left side
973 for (int i = 0; i < leftSpot; i++) { rightQuery[i] = '.'; }
978 catch(exception& e) {
979 m->errorOut(e, "ChimeraSlayerCommand", "divideInHalf");
983 /**************************************************************************************************/
984 map<string, int> ChimeraSlayerCommand::sortFastaFile(string fastaFile, string nameFile) {
986 map<string, int> nameAbund;
988 //read through fastafile and store info
989 map<string, string> seqs;
991 m->openInputFile(fastaFile, in);
995 if (m->control_pressed) { in.close(); return nameAbund; }
997 Sequence seq(in); m->gobble(in);
998 seqs[seq.getName()] = seq.getAligned();
1004 vector<seqPriorityNode> nameMapCount;
1005 int error = m->readNames(nameFile, nameMapCount, seqs);
1007 if (m->control_pressed) { return nameAbund; }
1009 if (error == 1) { m->control_pressed = true; return nameAbund; }
1010 if (seqs.size() != nameMapCount.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; }
1012 sort(nameMapCount.begin(), nameMapCount.end(), compareSeqPriorityNodes);
1014 string newFasta = fastaFile + ".temp";
1016 m->openOutputFile(newFasta, out);
1018 //print new file in order of
1019 for (int i = 0; i < nameMapCount.size(); i++) {
1020 out << ">" << nameMapCount[i].name << endl << nameMapCount[i].seq << endl;
1021 nameAbund[nameMapCount[i].name] = nameMapCount[i].numIdentical;
1025 rename(newFasta.c_str(), fastaFile.c_str());
1030 catch(exception& e) {
1031 m->errorOut(e, "ChimeraSlayerCommand", "sortFastaFile");
1036 /**************************************************************************************************/