vector<string> PreClusterCommand::setParameters(){
try {
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 pname("name", "InputTypes", "", "", "NameCount", "none", "none",false,false); parameters.push_back(pname);
+ CommandParameter pcount("count", "InputTypes", "", "", "NameCount-CountGroup", "none", "none",false,false); parameters.push_back(pcount);
+ CommandParameter pgroup("group", "InputTypes", "", "", "CountGroup", "none", "none",false,false); parameters.push_back(pgroup);
CommandParameter pdiffs("diffs", "Number", "", "0", "", "", "",false,false); parameters.push_back(pdiffs);
CommandParameter pprocessors("processors", "Number", "", "1", "", "", "",false,false); parameters.push_back(pprocessors);
CommandParameter pinputdir("inputdir", "String", "", "", "", "", "",false,false); parameters.push_back(pinputdir);
string helpString = "";
helpString += "The pre.cluster command groups sequences that are within a given number of base mismatches.\n";
helpString += "The pre.cluster command outputs a new fasta and name file.\n";
- helpString += "The pre.cluster command parameters are fasta, names and diffs. The fasta parameter is required. \n";
- helpString += "The names parameter allows you to give a list of seqs that are identical. This file is 2 columns, first column is name or representative sequence, second column is a list of its identical sequences separated by commas.\n";
+ helpString += "The pre.cluster command parameters are fasta, name, group, count, processors and diffs. The fasta parameter is required. \n";
+ helpString += "The name parameter allows you to give a list of seqs that are identical. This file is 2 columns, first column is name or representative sequence, second column is a list of its identical sequences separated by commas.\n";
helpString += "The group parameter allows you to provide a group file so you can cluster by group. \n";
+ helpString += "The count parameter allows you to provide a count file so you can cluster by group. \n";
helpString += "The diffs parameter allows you to specify maximum number of mismatched bases allowed between sequences in a grouping. The default is 1.\n";
helpString += "The pre.cluster command should be in the following format: \n";
helpString += "pre.cluster(fasta=yourFastaFile, names=yourNamesFile, diffs=yourMaxDiffs) \n";
else {
if (type == "fasta") { outputFileName = "precluster" + m->getExtension(inputName); }
else if (type == "name") { outputFileName = "precluster.names"; }
+ else if (type == "count") { outputFileName = "precluster.count_table"; }
else if (type == "map") { outputFileName = "precluster.map"; }
else { m->mothurOut("[ERROR]: No definition for type " + type + " output file tag.\n"); m->control_pressed = true; }
}
vector<string> tempOutNames;
outputTypes["fasta"] = tempOutNames;
outputTypes["name"] = tempOutNames;
+ outputTypes["count"] = tempOutNames;
outputTypes["map"] = tempOutNames;
}
catch(exception& e) {
outputTypes["fasta"] = tempOutNames;
outputTypes["name"] = tempOutNames;
outputTypes["map"] = tempOutNames;
+ outputTypes["count"] = tempOutNames;
//if the user changes the input directory command factory will send this info to us in the output parameter
string inputDir = validParameter.validFile(parameters, "inputdir", false);
//if the user has not given a path then, add inputdir. else leave path alone.
if (path == "") { parameters["group"] = inputDir + it->second; }
}
+
+ it = parameters.find("count");
+ //user has given a template file
+ if(it != parameters.end()){
+ path = m->hasPath(it->second);
+ //if the user has not given a path then, add inputdir. else leave path alone.
+ if (path == "") { parameters["count"] = inputDir + it->second; }
+ }
}
//check for required parameters
if (groupfile == "not found") { groupfile = ""; bygroup = false; }
else if (groupfile == "not open") { abort = true; groupfile = ""; }
else { m->setGroupFile(groupfile); bygroup = true; }
+
+ countfile = validParameter.validFile(parameters, "count", true);
+ if (countfile == "not found") { countfile = ""; }
+ else if (countfile == "not open") { abort = true; countfile = ""; }
+ else {
+ m->setCountTableFile(countfile);
+ ct.readTable(countfile);
+ if (ct.hasGroupInfo()) { bygroup = true; }
+ else { bygroup = false; }
+ }
+
+ if ((namefile != "") && (countfile != "")) {
+ m->mothurOut("[ERROR]: you may only use one of the following: name or count."); m->mothurOutEndLine(); abort = true;
+ }
+
+ if ((groupfile != "") && (countfile != "")) {
+ m->mothurOut("[ERROR]: you may only use one of the following: group or count."); m->mothurOutEndLine(); abort=true;
+ }
+
string temp = validParameter.validFile(parameters, "diffs", false); if(temp == "not found"){ temp = "1"; }
m->mothurConvert(temp, diffs);
m->setProcessors(temp);
m->mothurConvert(temp, processors);
- if (namefile == "") {
- vector<string> files; files.push_back(fastafile);
- parser.getNameFile(files);
- }
+ if (countfile == "") {
+ if (namefile == "") {
+ vector<string> files; files.push_back(fastafile);
+ parser.getNameFile(files);
+ }
+ }
}
}
string fileroot = outputDir + m->getRootName(m->getSimpleName(fastafile));
string newFastaFile = fileroot + getOutputFileNameTag("fasta", fastafile);
string newNamesFile = fileroot + getOutputFileNameTag("name");
+ string newCountFile = fileroot + getOutputFileNameTag("count");
string newMapFile = fileroot + getOutputFileNameTag("map"); //add group name if by group
outputNames.push_back(newFastaFile); outputTypes["fasta"].push_back(newFastaFile);
- outputNames.push_back(newNamesFile); outputTypes["name"].push_back(newNamesFile);
-
+ if (countfile == "") { outputNames.push_back(newNamesFile); outputTypes["name"].push_back(newNamesFile); }
+ else { outputNames.push_back(newCountFile); outputTypes["count"].push_back(newCountFile); }
if (bygroup) {
//clear out old files
newMapFile = fileroot + "precluster.";
//parse fasta and name file by group
- SequenceParser* parser;
- if (namefile != "") { parser = new SequenceParser(groupfile, fastafile, namefile); }
- else { parser = new SequenceParser(groupfile, fastafile); }
-
- vector<string> groups = parser->getNamesOfGroups();
-
- if(processors == 1) { driverGroups(parser, newFastaFile, newNamesFile, newMapFile, 0, groups.size(), groups); }
- else { createProcessesGroups(parser, newFastaFile, newNamesFile, newMapFile, groups); }
-
- delete parser;
-
- if (m->control_pressed) { for (int i = 0; i < outputNames.size(); i++) { m->mothurRemove(outputNames[i]); } return 0; }
-
- //run unique.seqs for deconvolute results
- string inputString = "fasta=" + newFastaFile;
- if (namefile != "") { inputString += ", name=" + newNamesFile; }
- m->mothurOutEndLine();
- m->mothurOut("/******************************************/"); m->mothurOutEndLine();
- m->mothurOut("Running command: unique.seqs(" + inputString + ")"); m->mothurOutEndLine();
- m->mothurCalling = true;
+ vector<string> groups;
+ if (countfile != "") {
+ cparser = new SequenceCountParser(countfile, fastafile);
+ groups = cparser->getNamesOfGroups();
+ }else {
+ if (namefile != "") { parser = new SequenceParser(groupfile, fastafile, namefile); }
+ else { parser = new SequenceParser(groupfile, fastafile); }
+ groups = parser->getNamesOfGroups();
+ }
- Command* uniqueCommand = new DeconvoluteCommand(inputString);
- uniqueCommand->execute();
-
- map<string, vector<string> > filenames = uniqueCommand->getOutputFiles();
-
- delete uniqueCommand;
- m->mothurCalling = false;
- m->mothurOut("/******************************************/"); m->mothurOutEndLine();
-
- m->renameFile(filenames["fasta"][0], newFastaFile);
- m->renameFile(filenames["name"][0], newNamesFile);
-
+ if(processors == 1) { driverGroups(newFastaFile, newNamesFile, newMapFile, 0, groups.size(), groups); }
+ else { createProcessesGroups(newFastaFile, newNamesFile, newMapFile, groups); }
+
+ if (countfile != "") {
+ mergeGroupCounts(newCountFile, newNamesFile, newFastaFile);
+ delete cparser;
+ }else {
+ delete parser;
+ //run unique.seqs for deconvolute results
+ string inputString = "fasta=" + newFastaFile;
+ if (namefile != "") { inputString += ", name=" + newNamesFile; }
+ m->mothurOutEndLine();
+ m->mothurOut("/******************************************/"); m->mothurOutEndLine();
+ m->mothurOut("Running command: unique.seqs(" + inputString + ")"); m->mothurOutEndLine();
+ m->mothurCalling = true;
+
+ Command* uniqueCommand = new DeconvoluteCommand(inputString);
+ uniqueCommand->execute();
+
+ map<string, vector<string> > filenames = uniqueCommand->getOutputFiles();
+
+ delete uniqueCommand;
+ m->mothurCalling = false;
+ m->mothurOut("/******************************************/"); m->mothurOutEndLine();
+
+ m->renameFile(filenames["fasta"][0], newFastaFile);
+ m->renameFile(filenames["name"][0], newNamesFile);
+ }
+ if (m->control_pressed) { for (int i = 0; i < outputNames.size(); i++) { m->mothurRemove(outputNames[i]); } return 0; }
m->mothurOut("It took " + toString(time(NULL) - start) + " secs to run pre.cluster."); m->mothurOutEndLine();
}else {
m->mothurOut("Total number of sequences before precluster was " + toString(alignSeqs.size()) + "."); m->mothurOutEndLine();
m->mothurOut("pre.cluster removed " + toString(count) + " sequences."); m->mothurOutEndLine(); m->mothurOutEndLine();
- printData(newFastaFile, newNamesFile);
-
+ if (countfile != "") { newNamesFile = newCountFile; }
+ printData(newFastaFile, newNamesFile, "");
+
m->mothurOut("It took " + toString(time(NULL) - start) + " secs to cluster " + toString(numSeqs) + " sequences."); m->mothurOutEndLine();
}
if (itTypes != outputTypes.end()) {
if ((itTypes->second).size() != 0) { current = (itTypes->second)[0]; m->setNameFile(current); }
}
+
+ itTypes = outputTypes.find("count");
+ if (itTypes != outputTypes.end()) {
+ if ((itTypes->second).size() != 0) { current = (itTypes->second)[0]; m->setCountTableFile(current); }
+ }
return 0;
}
}
/**************************************************************************************************/
-int PreClusterCommand::createProcessesGroups(SequenceParser* parser, string newFName, string newNName, string newMFile, vector<string> groups) {
+int PreClusterCommand::createProcessesGroups(string newFName, string newNName, string newMFile, vector<string> groups) {
try {
vector<int> processIDS;
process++;
}else if (pid == 0){
outputNames.clear();
- num = driverGroups(parser, newFName + toString(getpid()) + ".temp", newNName + toString(getpid()) + ".temp", newMFile, lines[process].start, lines[process].end, groups);
+ num = driverGroups(newFName + toString(getpid()) + ".temp", newNName + toString(getpid()) + ".temp", newMFile, lines[process].start, lines[process].end, groups);
string tempFile = toString(getpid()) + ".outputNames.temp";
ofstream outTemp;
}
//do my part
- num = driverGroups(parser, newFName, newNName, newMFile, lines[0].start, lines[0].end, groups);
+ num = driverGroups(newFName, newNName, newMFile, lines[0].start, lines[0].end, groups);
//force parent to wait until all the processes are done
for (int i=0;i<processIDS.size();i++) {
// Allocate memory for thread data.
string extension = toString(i) + ".temp";
- preClusterData* tempPreCluster = new preClusterData(fastafile, namefile, groupfile, (newFName+extension), (newNName+extension), newMFile, groups, m, lines[i].start, lines[i].end, diffs, i);
+ preClusterData* tempPreCluster = new preClusterData(fastafile, namefile, groupfile, countfile, (newFName+extension), (newNName+extension), newMFile, groups, m, lines[i].start, lines[i].end, diffs, i);
pDataArray.push_back(tempPreCluster);
processIDS.push_back(i);
//using the main process as a worker saves time and memory
- num = driverGroups(parser, newFName, newNName, newMFile, lines[0].start, lines[0].end, groups);
+ num = driverGroups(newFName, newNName, newMFile, lines[0].start, lines[0].end, groups);
//Wait until all threads have terminated.
WaitForMultipleObjects(processors-1, hThreadArray, TRUE, INFINITE);
}
}
/**************************************************************************************************/
-int PreClusterCommand::driverGroups(SequenceParser* parser, string newFFile, string newNFile, string newMFile, int start, int end, vector<string> groups){
+int PreClusterCommand::driverGroups(string newFFile, string newNFile, string newMFile, int start, int end, vector<string> groups){
try {
int numSeqs = 0;
m->mothurOutEndLine(); m->mothurOut("Processing group " + groups[i] + ":"); m->mothurOutEndLine();
map<string, string> thisNameMap;
- if (namefile != "") { thisNameMap = parser->getNameMap(groups[i]); }
- vector<Sequence> thisSeqs = parser->getSeqs(groups[i]);
-
+ vector<Sequence> thisSeqs;
+ if (groupfile != "") {
+ thisSeqs = parser->getSeqs(groups[i]);
+ }else if (countfile != "") {
+ thisSeqs = cparser->getSeqs(groups[i]);
+ }
+ if (namefile != "") { thisNameMap = parser->getNameMap(groups[i]); }
+
//fill alignSeqs with this groups info.
- numSeqs = loadSeqs(thisNameMap, thisSeqs);
+ numSeqs = loadSeqs(thisNameMap, thisSeqs, groups[i]);
if (m->control_pressed) { return 0; }
if (diffs > length) { m->mothurOut("Error: diffs is greater than your sequence length."); m->mothurOutEndLine(); m->control_pressed = true; return 0; }
- int count = process(newMFile+groups[i]+".map");
+ int count= process(newMFile+groups[i]+".map");
outputNames.push_back(newMFile+groups[i]+".map"); outputTypes["map"].push_back(newMFile+groups[i]+".map");
if (m->control_pressed) { return 0; }
m->mothurOut("Total number of sequences before pre.cluster was " + toString(alignSeqs.size()) + "."); m->mothurOutEndLine();
m->mothurOut("pre.cluster removed " + toString(count) + " sequences."); m->mothurOutEndLine(); m->mothurOutEndLine();
- printData(newFFile, newNFile);
+ printData(newFFile, newNFile, groups[i]);
m->mothurOut("It took " + toString(time(NULL) - start) + " secs to cluster " + toString(numSeqs) + " sequences."); m->mothurOutEndLine();
//ifstream inNames;
ifstream inFasta;
- //m->openInputFile(namefile, inNames);
m->openInputFile(fastafile, inFasta);
-
- //string firstCol, secondCol, nameString;
set<int> lengths;
while (!inFasta.eof()) {
if (m->control_pressed) { inFasta.close(); return 0; }
-
- //inNames >> firstCol >> secondCol;
- //nameString = secondCol;
-
- //m->gobble(inNames);
- //int size = 1;
- //while (secondCol.find_first_of(',') != -1) {
- // size++;
- // secondCol = secondCol.substr(secondCol.find_first_of(',')+1, secondCol.length());
- //}
-
+
Sequence seq(inFasta); m->gobble(inFasta);
if (seq.getName() != "") { //can get "" if commented line is at end of fasta file
lengths.insert(seq.getAligned().length());
}
}else { //no names file, you are identical to yourself
- seqPNode tempNode(1, seq, seq.getName());
+ int numRep = 1;
+ if (countfile != "") { numRep = ct.getNumSeqs(seq.getName()); }
+ seqPNode tempNode(numRep, seq, seq.getName());
alignSeqs.push_back(tempNode);
lengths.insert(seq.getAligned().length());
}
}
}
inFasta.close();
- //inNames.close();
if (lengths.size() > 1) { m->control_pressed = true; m->mothurOut("[ERROR]: your sequences are not all the same length. pre.cluster requires sequences to be aligned."); m->mothurOutEndLine(); }
else if (lengths.size() == 1) { length = *(lengths.begin()); }
}
}
/**************************************************************************************************/
-int PreClusterCommand::loadSeqs(map<string, string>& thisName, vector<Sequence>& thisSeqs){
+int PreClusterCommand::loadSeqs(map<string, string>& thisName, vector<Sequence>& thisSeqs, string group){
try {
set<int> lengths;
alignSeqs.clear();
map<string, string>::iterator it;
bool error = false;
-
+ map<string, int> thisCount;
+ if (countfile != "") { thisCount = cparser->getCountTable(group); }
+
for (int i = 0; i < thisSeqs.size(); i++) {
if (m->control_pressed) { return 0; }
lengths.insert(thisSeqs[i].getAligned().length());
}
}else { //no names file, you are identical to yourself
- seqPNode tempNode(1, thisSeqs[i], thisSeqs[i].getName());
+ int numRep = 1;
+ if (countfile != "") {
+ map<string, int>::iterator it2 = thisCount.find(thisSeqs[i].getName());
+
+ //should never be true since parser checks for this
+ if (it2 == thisCount.end()) { m->mothurOut(thisSeqs[i].getName() + " is not in your count file, please correct."); m->mothurOutEndLine(); error = true; }
+ else { numRep = it2->second; }
+ }
+ seqPNode tempNode(numRep, thisSeqs[i], thisSeqs[i].getName());
alignSeqs.push_back(tempNode);
lengths.insert(thisSeqs[i].getAligned().length());
}
}
-
+
if (lengths.size() > 1) { error = true; m->mothurOut("[ERROR]: your sequences are not all the same length. pre.cluster requires sequences to be aligned."); m->mothurOutEndLine(); }
else if (lengths.size() == 1) { length = *(lengths.begin()); }
exit(1);
}
}
+/**************************************************************************************************/
+
+int PreClusterCommand::mergeGroupCounts(string newcount, string newname, string newfasta){
+ try {
+ ifstream inNames;
+ m->openInputFile(newname, inNames);
+
+ string group, first, second;
+ set<string> uniqueNames;
+ while (!inNames.eof()) {
+ if (m->control_pressed) { break; }
+ inNames >> group; m->gobble(inNames);
+ inNames >> first; m->gobble(inNames);
+ inNames >> second; m->gobble(inNames);
+
+ vector<string> names;
+ m->splitAtComma(second, names);
+
+ uniqueNames.insert(first);
+
+ int total = ct.getGroupCount(first, group);
+ for (int i = 1; i < names.size(); i++) {
+ total += ct.getGroupCount(names[i], group);
+ ct.setAbund(names[i], group, 0);
+ }
+ ct.setAbund(first, group, total);
+ }
+ inNames.close();
+
+ vector<string> namesOfSeqs = ct.getNamesOfSeqs();
+ for (int i = 0; i < namesOfSeqs.size(); i++) {
+ if (ct.getNumSeqs(namesOfSeqs[i]) == 0) {
+ ct.remove(namesOfSeqs[i]);
+ }
+ }
+
+ ct.printTable(newcount);
+ m->mothurRemove(newname);
+
+ if (bygroup) { //if by group, must remove the duplicate seqs that are named the same
+ ifstream in;
+ m->openInputFile(newfasta, in);
+
+ ofstream out;
+ m->openOutputFile(newfasta+"temp", out);
+
+ int count = 0;
+ set<string> already;
+ while(!in.eof()) {
+ if (m->control_pressed) { break; }
+
+ Sequence seq(in); m->gobble(in);
+
+ if (seq.getName() != "") {
+ count++;
+ if (already.count(seq.getName()) == 0) {
+ seq.printSequence(out);
+ already.insert(seq.getName());
+ }
+ }
+ }
+ in.close();
+ out.close();
+ m->mothurRemove(newfasta);
+ m->renameFile(newfasta+"temp", newfasta);
+ }
+ return 0;
+
+ }
+ catch(exception& e) {
+ m->errorOut(e, "PreClusterCommand", "mergeGroupCounts");
+ exit(1);
+ }
+}
/**************************************************************************************************/
-void PreClusterCommand::printData(string newfasta, string newname){
+void PreClusterCommand::printData(string newfasta, string newname, string group){
try {
ofstream outFasta;
ofstream outNames;
m->openOutputFile(newname, outNames);
}
+ if ((countfile != "") && (group == "")) { outNames << "Representative_Sequence\ttotal\n"; }
for (int i = 0; i < alignSeqs.size(); i++) {
if (alignSeqs[i].numIdentical != 0) {
alignSeqs[i].seq.printSequence(outFasta);
- outNames << alignSeqs[i].seq.getName() << '\t' << alignSeqs[i].names << endl;
+ if (countfile != "") {
+ if (group != "") { outNames << group << '\t' << alignSeqs[i].seq.getName() << '\t' << alignSeqs[i].names << endl; }
+ else { outNames << alignSeqs[i].seq.getName() << '\t' << alignSeqs[i].numIdentical << endl; }
+ }else { outNames << alignSeqs[i].seq.getName() << '\t' << alignSeqs[i].names << endl; }
}
}
projects / mothur.git / blobdiff