#include "filterseqscommand.h"
/**************************************************************************************/
-void FilterSeqsCommand::doTrump() {
- trump = globaldata->getTrump();
- for(int i = 0; i < db->size(); i++) {
- Sequence cur = db->get(i);
- string curAligned = cur.getUnaligned();
- for(int j = 0; j < curAligned.length(); j++) {
- string curChar = curAligned.substr(j, 1);
- if(curChar.compare(trump) == 0)
- columnsToRemove[j] = true;
- }
- }
+
+FilterSeqsCommand::FilterSeqsCommand(){
+
+ globaldata = GlobalData::getInstance();
+
+ if(globaldata->getFastaFile() == "") { cout << "You must enter a fasta formatted file" << endl; }
+ trump = globaldata->getTrump()[0];
+ numSeqs = 0;
+
}
/**************************************************************************************/
-void FilterSeqsCommand::doSoft() {
- soft = atoi(globaldata->getSoft().c_str());
- vector<vector<int> > columnSymbolSums;
- vector<vector<string> > columnSymbols;
- for(int i = 0; i < db->get(0).getLength(); i++) {
- vector<string> symbols;
- vector<int> sums;
- columnSymbols.push_back(symbols);
- columnSymbolSums.push_back(sums);
- }
+
+void FilterSeqsCommand::doHard() {
- for(int i = 0; i < db->size(); i++) {
- Sequence cur = db->get(i);
- string curAligned = cur.getUnaligned();
+ string hardName = globaldata->getHard();
+ string hardFilter = "";
- for(int j = 0; j < curAligned.length(); j++) {
- string curChar = curAligned.substr(j, 1);
- vector<string> curColumnSymbols = columnSymbols[j];
- bool newSymbol = true;
-
- for(int k = 0; k < curColumnSymbols.size(); k++)
- if(curChar.compare(curColumnSymbols[k]) == 0) {
- newSymbol = false;
- columnSymbolSums[j][k]++;
- }
-
- if(newSymbol) {
- columnSymbols[j].push_back(curChar);
- columnSymbolSums[j].push_back(1);
- }
+ ifstream fileHandle;
+ openInputFile(hardName, fileHandle);
+
+ fileHandle >> filter;
+
+}
+
+/**************************************************************************************/
+
+void FilterSeqsCommand::doTrump(Sequence seq) {
+
+ string curAligned = seq.getAligned();
+
+ for(int j = 0; j < alignmentLength; j++) {
+ if(curAligned[j] == trump){
+ filter[j] = '0';
}
}
+
+}
+
+/**************************************************************************************/
+
+void FilterSeqsCommand::doVertical() {
+
+ for(int i=0;i<alignmentLength;i++){
+ if(gap[i] == numSeqs) { filter[i] = '0'; }
+ }
+
+}
+
+/**************************************************************************************/
+
+void FilterSeqsCommand::doSoft() {
+ int threshold = int (soft * numSeqs);
+ bool keep = 0;
- for(int i = 0; i < columnSymbolSums.size(); i++) {
- int totalSum = 0;
- int max = 0;
- vector<int> curColumnSymbols = columnSymbolSums[i];
-
- for(int j = 0; j < curColumnSymbols.size(); j++) {
- int curSum = curColumnSymbols[j];
- //cout << columnSymbols[i][j] << ": " << curSum << "\n";
- if(curSum > max)
- max = curSum;
- totalSum += curSum;
- }
- //cout << "\n";
+ for(int i=0;i<alignmentLength;i++){
+ if(a[i] >= threshold) { keep = 1; }
+ else if(t[i] >= threshold) { keep = 1; }
+ else if(g[i] >= threshold) { keep = 1; }
+ else if(c[i] >= threshold) { keep = 1; }
- if((double)max/(double)totalSum * 100 < soft)
- columnsToRemove[i] = true;
+ if(keep == 0) { filter[i] = 0; }
}
}
/**************************************************************************************/
-void FilterSeqsCommand::doFilter() {
- filter = globaldata->getFilter();
- ifstream filehandle;
- openInputFile(filter, filehandle);
+
+void FilterSeqsCommand::getFreqs(Sequence seq) {
+
+ string curAligned = seq.getAligned();;
- char c;
- int count = 0;
- while(!filehandle.eof()) {
- c = filehandle.get();
- if(c == '0')
- columnsToRemove[count] = true;
- count++;
+ for(int j=0;j<alignmentLength;j++){
+ if(toupper(curAligned[j]) == 'A') { a[j]++; }
+ else if(toupper(curAligned[j]) == 'T' || toupper(curAligned[j]) == 'U') { t[j]++; }
+ else if(toupper(curAligned[j]) == 'G') { g[j]++; }
+ else if(toupper(curAligned[j]) == 'C') { c[j]++; }
+ else if(curAligned[j] == '-' || curAligned[j] == '.') { gap[j]++; }
}
+
}
/**************************************************************************************/
+
int FilterSeqsCommand::execute() {
try {
- globaldata = GlobalData::getInstance();
- filename = globaldata->inputFileName;
+ ifstream inFASTA;
+ openInputFile(globaldata->getFastaFile(), inFASTA);
+
+ Sequence testSeq(inFASTA);
+ alignmentLength = testSeq.getAlignLength();
+ inFASTA.seekg(0);
+
+ if(globaldata->getSoft() != "" || isTrue(globaldata->getVertical())){
+ a.assign(alignmentLength, 0);
+ t.assign(alignmentLength, 0);
+ g.assign(alignmentLength, 0);
+ c.assign(alignmentLength, 0);
+ gap.assign(alignmentLength, 0);
+ }
+ if(globaldata->getSoft() != ""){
+ soft = (float)atoi(globaldata->getSoft().c_str()) / 100.0;
+ }
+
+ if(globaldata->getHard().compare("") != 0) { doHard(); }
+ else { filter = string(alignmentLength, '1'); }
+
+ if(globaldata->getTrump().compare("") != 0 || isTrue(globaldata->getVertical()) || globaldata->getSoft().compare("") != 0){
+
+ while(!inFASTA.eof()){
+ Sequence seq(inFASTA);
+ if(globaldata->getTrump().compare("") != 0) { doTrump(seq); }
+ if(isTrue(globaldata->getVertical()) || globaldata->getSoft().compare("") != 0){ getFreqs(seq); }
+ numSeqs++;
+ cout.flush();
+ }
- if(globaldata->getFastaFile() != "") {
- readSeqs = new ReadFasta(filename); }
- else if(globaldata->getNexusFile() != "") {
- readSeqs = new ReadNexus(filename); }
- else if(globaldata->getClustalFile() != "") {
- readSeqs = new ReadClustal(filename); }
- else if(globaldata->getPhylipFile() != "") {
- readSeqs = new ReadPhylip(filename); }
-
- readSeqs->read();
- db = readSeqs->getDB();
+ }
+ inFASTA.close();
- //for(int i = 0; i < db->size(); i++) {
-// cout << db->get(i).getLength() << "\n" << db->get(i).getName() << ": " << db->get(i).getUnaligned() << "\n\n";
-// }
+ if(isTrue(globaldata->getVertical()) == 1) { doVertical(); }
+ if(globaldata->getSoft().compare("") != 0) { doSoft(); }
- for(int i = 0; i < db->get(0).getLength(); i++)
- columnsToRemove.push_back(false);
+ ofstream outFilter;
+ string filterFile = getRootName(globaldata->inputFileName) + "filter";
+ openOutputFile(filterFile, outFilter);
+ outFilter << filter << endl;
+ outFilter.close();
+
+
+ openInputFile(globaldata->getFastaFile(), inFASTA);
+ string filteredFasta = getRootName(globaldata->inputFileName) + "filter.fasta";
+ ofstream outFASTA;
+ openOutputFile(filteredFasta, outFASTA);
+
+ numSeqs = 0;
+ while(!inFASTA.eof()){
+ Sequence seq(inFASTA);
+ string align = seq.getAligned();
+ string filterSeq = "";
+
+ for(int j=0;j<alignmentLength;j++){
+ if(filter[j] == '1'){
+ filterSeq += align[j];
+ }
+ }
+
+ outFASTA << '>' << seq.getName() << endl << filterSeq << endl;
+ numSeqs++;
+ gobble(inFASTA);
+ }
+ outFASTA.close();
+ inFASTA.close();
-
- if(globaldata->getTrump().compare("") != 0)
- doTrump();
- else if(globaldata->getSoft().compare("") != 0)
- doSoft();
- else if(globaldata->getFilter().compare("") != 0)
- doFilter();
- //for(int i = 0; i < columnsToRemove.size(); i++)
-// {
-// cout << "Remove Column " << i << " = ";
-// if(columnsToRemove[i])
-// cout << "true\n";
-// else
-// cout << "false\n";
-// }
-
-
- //Creating the new SequenceDB
- SequenceDB newDB;
- for(int i = 0; i < db->size(); i++) {
- Sequence curSeq = db->get(i);
- string curAligned = curSeq.getUnaligned();
- string curName = curSeq.getName();
- string newAligned = "";
- for(int j = 0; j < curAligned.length(); j++)
- if(!columnsToRemove[j])
- newAligned += curAligned.substr(j, 1);
-
- Sequence newSeq(curName, newAligned);
- newDB.add(newSeq);
+ int filteredLength = 0;
+ for(int i=0;i<alignmentLength;i++){
+ if(filter[i] == '1'){ filteredLength++; }
}
- string newFileName = getRootName(filename) + "filter.fa";
- ofstream outfile;
- outfile.open(newFileName.c_str());
- newDB.print(outfile);
- outfile.close();
-
+ cout << endl;
+ cout << "Length of filtered alignment: " << filteredLength << endl;
+ cout << "Number of columns removed: " << alignmentLength-filteredLength << endl;
+ cout << "Length of the original alignment: " << alignmentLength << endl;
+ cout << "Number of sequences used to construct filter: " << numSeqs << endl;
+
globaldata->clear();
- //delete db;
- //delete newDB;
return 0;
+
}
catch(exception& e) {
cout << "Standard Error: " << e.what() << " has occurred in the FilterSeqsCommand class Function execute. Please contact Pat Schloss at pschloss@microbio.umass.edu." << "\n";
cout << "An unknown error has occurred in the FilterSeqsCommand class function execute. Please contact Pat Schloss at pschloss@microbio.umass.edu." << "\n";
exit(1);
}
-}
\ No newline at end of file
+}
+
+/**************************************************************************************/