//**********************************************************************************************************************
RemoveLineageCommand::RemoveLineageCommand(){
try {
+ abort = true;
//initialize outputTypes
vector<string> tempOutNames;
outputTypes["fasta"] = tempOutNames;
else if (taxfile == "not found") { taxfile = ""; m->mothurOut("The taxonomy parameter is required for the get.lineage command."); m->mothurOutEndLine(); abort = true; }
string usedDups = "true";
- string temp = validParameter.validFile(parameters, "dups", false); if (temp == "not found") { temp = "false"; usedDups = ""; }
+ string temp = validParameter.validFile(parameters, "dups", false);
+ if (temp == "not found") {
+ if (namefile != "") { temp = "true"; }
+ else { temp = "false"; usedDups = ""; }
+ }
dups = m->isTrue(temp);
taxons = validParameter.validFile(parameters, "taxon", false);
m->mothurOut("The remove.lineage command parameters are taxon, fasta, name, group, list, taxonomy, alignreport and dups. You must provide taxonomy and taxon.\n");
m->mothurOut("The dups parameter allows you to add the entire line from a name file if you add any name from the line. default=false. \n");
m->mothurOut("The taxon parameter allows you to select the taxons you would like to remove.\n");
+ m->mothurOut("You may enter your taxons with confidence scores, doing so will remove only those sequences that belong to the taxonomy and whose cofidence scores fall below the scores you give.\n");
+ m->mothurOut("If they belong to the taxonomy and have confidences above those you provide the sequence will not be removed.\n");
m->mothurOut("The remove.lineage command should be in the following format: remove.lineage(taxonomy=yourTaxonomyFile, taxon=yourTaxons).\n");
m->mothurOut("Example remove.lineage(taxonomy=amazon.silva.taxonomy, taxon=Bacteria;Firmicutes;Bacilli;Lactobacillales;).\n");
m->mothurOut("Note: If you are running mothur in script mode you must wrap the taxon in ' characters so mothur will ignore the ; in the taxon.\n");
bool wroteSomething = false;
+ bool taxonsHasConfidence = false;
+ vector< map<string, int> > searchTaxons;
+ string noConfidenceTaxons = taxons;
+ int hasConPos = taxons.find_first_of('(');
+ if (hasConPos != string::npos) {
+ taxonsHasConfidence = true;
+ searchTaxons = getTaxons(taxons);
+ noConfidenceTaxons = removeConfidences(taxons);
+ }
+
+
while(!in.eof()){
if (m->control_pressed) { in.close(); out.close(); remove(outputFileName.c_str()); return 0; }
string newtax = tax;
- //if the users file contains confidence scores we want to ignore them when searching for the taxons
- int hasConfidences = tax.find_first_of('(');
- if (hasConfidences != string::npos) {
- newtax = removeConfidences(tax);
+ //if the users file contains confidence scores we want to ignore them when searching for the taxons, unless the taxon has them
+ if (!taxonsHasConfidence) {
+
+ int hasConfidences = tax.find_first_of('(');
+ if (hasConfidences != string::npos) {
+ newtax = removeConfidences(tax);
+ }
+
+ int pos = newtax.find(taxons);
+
+ if (pos == string::npos) {
+ wroteSomething = true;
+ out << name << '\t' << tax << endl;
+ }else{ //this sequence contains the taxon the user wants to remove
+ names.insert(name);
+ }
+
+ }else{//if taxons has them and you don't them remove taxons
+ int hasConfidences = tax.find_first_of('(');
+ if (hasConfidences == string::npos) {
+
+ int pos = newtax.find(noConfidenceTaxons);
+
+ if (pos == string::npos) {
+ wroteSomething = true;
+ out << name << '\t' << tax << endl;
+ }else{ //this sequence contains the taxon the user wants to remove
+ names.insert(name);
+ }
+ }else { //both have confidences so we want to make sure the users confidences are greater then or equal to the taxons
+ //first remove confidences from both and see if the taxonomy exists
+
+ string noNewTax = tax;
+ int hasConfidences = tax.find_first_of('(');
+ if (hasConfidences != string::npos) {
+ noNewTax = removeConfidences(tax);
+ }
+
+ int pos = noNewTax.find(noConfidenceTaxons);
+
+ if (pos != string::npos) { //if yes, then are the confidences okay
+
+ bool remove = false;
+ vector< map<string, int> > usersTaxon = getTaxons(newtax);
+
+ //the usersTaxon is most likely longer than the searchTaxons, and searchTaxon[0] may relate to userTaxon[4]
+ //we want to "line them up", so we will find the the index where the searchstring starts
+ int index = 0;
+ for (int i = 0; i < usersTaxon.size(); i++) {
+
+ if (usersTaxon[i].begin()->first == searchTaxons[0].begin()->first) {
+ index = i;
+ int spot = 0;
+ bool goodspot = true;
+ //is this really the start, or are we dealing with a taxon of the same name?
+ while ((spot < searchTaxons.size()) && ((i+spot) < usersTaxon.size())) {
+ if (usersTaxon[i+spot].begin()->first != searchTaxons[spot].begin()->first) { goodspot = false; break; }
+ else { spot++; }
+ }
+
+ if (goodspot) { break; }
+ }
+ }
+
+ for (int i = 0; i < searchTaxons.size(); i++) {
+
+ if ((i+index) < usersTaxon.size()) { //just in case, should never be false
+ if (usersTaxon[i+index].begin()->second < searchTaxons[i].begin()->second) { //is the users cutoff less than the search taxons
+ remove = true;
+ break;
+ }
+ }else {
+ remove = true;
+ break;
+ }
+ }
+
+ //passed the test so remove you
+ if (remove) {
+ names.insert(name);
+ }else {
+ wroteSomething = true;
+ out << name << '\t' << tax << endl;
+ }
+ }else {
+ wroteSomething = true;
+ out << name << '\t' << tax << endl;
+ }
+ }
}
- int pos = newtax.find(taxons);
- if (pos == string::npos) {
- wroteSomething = true;
- out << name << '\t' << tax << endl;
- }else{ //this sequence contains the taxon the user wants to remove
- names.insert(name);
- }
+
m->gobble(in);
}
}
}
/**************************************************************************************************/
+vector< map<string, int> > RemoveLineageCommand::getTaxons(string tax) {
+ try {
+
+ vector< map<string, int> > t;
+ string taxon = "";
+ int taxLength = tax.length();
+ for(int i=0;i<taxLength;i++){
+ if(tax[i] == ';'){
+
+ int openParen = taxon.find_first_of('(');
+ int closeParen = taxon.find_last_of(')');
+
+ string newtaxon, confidence;
+ if ((openParen != string::npos) && (closeParen != string::npos)) {
+ newtaxon = taxon.substr(0, openParen); //rip off confidence
+ confidence = taxon.substr((openParen+1), (closeParen-openParen-1));
+ }else{
+ newtaxon = taxon;
+ confidence = "0";
+ }
+ int con = 0;
+ convert(confidence, con);
+
+ map<string, int> temp;
+ temp[newtaxon] = con;
+ t.push_back(temp);
+
+ taxon = "";
+ }
+ else{
+ taxon += tax[i];
+ }
+ }
+
+ return t;
+ }
+ catch(exception& e) {
+ m->errorOut(e, "RemoveLineageCommand", "getTaxons");
+ exit(1);
+ }
+}
+/**************************************************************************************************/
string RemoveLineageCommand::removeConfidences(string tax) {
try {