CommandParameter pname("name", "InputTypes", "", "", "NameCount", "none", "none","name",false,false,true); parameters.push_back(pname);
CommandParameter pcount("count", "InputTypes", "", "", "NameCount-CountGroup", "none", "none","count",false,false,true); parameters.push_back(pcount);
CommandParameter pgroup("group", "InputTypes", "", "", "CountGroup", "none", "none","",false,false,true); parameters.push_back(pgroup);
- CommandParameter pdiffs("diffs", "Number", "", "0", "", "", "","",false,false,true); parameters.push_back(pdiffs);
+ CommandParameter pdiffs("diffs", "Number", "", "1", "", "", "","",false,false,true); parameters.push_back(pdiffs);
CommandParameter pprocessors("processors", "Number", "", "1", "", "", "","",false,false,true); parameters.push_back(pprocessors);
+ CommandParameter ptopdown("topdown", "Boolean", "", "T", "", "", "","",false,false); parameters.push_back(ptopdown);
+
CommandParameter pinputdir("inputdir", "String", "", "", "", "", "","",false,false); parameters.push_back(pinputdir);
CommandParameter poutputdir("outputdir", "String", "", "", "", "", "","",false,false); parameters.push_back(poutputdir);
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, name, group, count, processors and diffs. The fasta parameter is required. \n";
+ helpString += "The pre.cluster command parameters are fasta, name, group, count, topdown, 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 topdown parameter allows you to specify whether to cluster from largest abundance to smallest or smallest to largest. Default=T, meaning largest to smallest.\n";
helpString += "The pre.cluster command should be in the following format: \n";
helpString += "pre.cluster(fasta=yourFastaFile, names=yourNamesFile, diffs=yourMaxDiffs) \n";
helpString += "Example pre.cluster(fasta=amazon.fasta, diffs=2).\n";
else if (countfile == "not open") { abort = true; countfile = ""; }
else {
m->setCountTableFile(countfile);
- ct.readTable(countfile);
+ ct.readTable(countfile, true, false);
if (ct.hasGroupInfo()) { bygroup = true; }
else { bygroup = false; }
}
m->setProcessors(temp);
m->mothurConvert(temp, processors);
+ temp = validParameter.validFile(parameters, "topdown", false); if(temp == "not found"){ temp = "T"; }
+ topdown = m->isTrue(temp);
+
if (countfile == "") {
if (namefile == "") {
vector<string> files; files.push_back(fastafile);
m->mothurOut("It took " + toString(time(NULL) - start) + " secs to run pre.cluster."); m->mothurOutEndLine();
}else {
+ if (processors != 1) { m->mothurOut("When using running without group information mothur can only use 1 processor, continuing."); m->mothurOutEndLine(); processors = 1; }
if (namefile != "") { readNameFile(); }
//reads fasta file and return number of seqs
//divide the groups between the processors
vector<linePair> lines;
- int numGroupsPerProcessor = groups.size() / processors;
- for (int i = 0; i < processors; i++) {
- int startIndex = i * numGroupsPerProcessor;
- int endIndex = (i+1) * numGroupsPerProcessor;
- if(i == (processors - 1)){ endIndex = groups.size(); }
- lines.push_back(linePair(startIndex, endIndex));
- }
+ int remainingPairs = groups.size();
+ int startIndex = 0;
+ for (int remainingProcessors = processors; remainingProcessors > 0; remainingProcessors--) {
+ int numPairs = remainingPairs; //case for last processor
+ if (remainingProcessors != 1) { numPairs = ceil(remainingPairs / remainingProcessors); }
+ lines.push_back(linePair(startIndex, (startIndex+numPairs))); //startIndex, endIndex
+ startIndex = startIndex + numPairs;
+ remainingPairs = remainingPairs - numPairs;
+ }
#if defined (__APPLE__) || (__MACH__) || (linux) || (__linux) || (__linux__) || (__unix__) || (__unix)
// Allocate memory for thread data.
string extension = toString(i) + ".temp";
- preClusterData* tempPreCluster = new preClusterData(fastafile, namefile, groupfile, countfile, (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, topdown, i);
pDataArray.push_back(tempPreCluster);
processIDS.push_back(i);
//Close all thread handles and free memory allocations.
for(int i=0; i < pDataArray.size(); i++){
+ if (pDataArray[i]->count != (pDataArray[i]->end-pDataArray[i]->start)) {
+ m->mothurOut("[ERROR]: process " + toString(i) + " only processed " + toString(pDataArray[i]->count) + " of " + toString(pDataArray[i]->end-pDataArray[i]->start) + " groups assigned to it, quitting. \n"); m->control_pressed = true;
+ }
for (int j = 0; j < pDataArray[i]->mapFileNames.size(); j++) {
outputNames.push_back(pDataArray[i]->mapFileNames[j]); outputTypes["map"].push_back(pDataArray[i]->mapFileNames[j]);
}
m->openOutputFile(newMapFile, out);
//sort seqs by number of identical seqs
- sort(alignSeqs.begin(), alignSeqs.end(), comparePriority);
+ if (topdown) { sort(alignSeqs.begin(), alignSeqs.end(), comparePriorityTopDown); }
+ else { sort(alignSeqs.begin(), alignSeqs.end(), comparePriorityDownTop); }
int count = 0;
int numSeqs = alignSeqs.size();
- //think about running through twice...
- for (int i = 0; i < numSeqs; i++) {
-
- //are you active
- // itActive = active.find(alignSeqs[i].seq.getName());
-
- if (alignSeqs[i].active) { //this sequence has not been merged yet
-
- string chunk = alignSeqs[i].seq.getName() + "\t" + toString(alignSeqs[i].numIdentical) + "\t" + toString(0) + "\t" + alignSeqs[i].seq.getAligned() + "\n";
-
- //try to merge it with all smaller seqs
- for (int j = i+1; j < numSeqs; j++) {
-
- if (m->control_pressed) { out.close(); return 0; }
-
- if (alignSeqs[j].active) { //this sequence has not been merged yet
- //are you within "diff" bases
- int mismatch = calcMisMatches(alignSeqs[i].seq.getAligned(), alignSeqs[j].seq.getAligned());
-
- if (mismatch <= diffs) {
- //merge
- alignSeqs[i].names += ',' + alignSeqs[j].names;
- alignSeqs[i].numIdentical += alignSeqs[j].numIdentical;
-
- chunk += alignSeqs[j].seq.getName() + "\t" + toString(alignSeqs[j].numIdentical) + "\t" + toString(mismatch) + "\t" + alignSeqs[j].seq.getAligned() + "\n";
-
- alignSeqs[j].active = 0;
- alignSeqs[j].numIdentical = 0;
- count++;
- }
- }//end if j active
- }//end for loop j
-
- //remove from active list
- alignSeqs[i].active = 0;
-
- out << "ideal_seq_" << (i+1) << '\t' << alignSeqs[i].numIdentical << endl << chunk << endl;;
-
- }//end if active i
- if(i % 100 == 0) { m->mothurOut(toString(i) + "\t" + toString(numSeqs - count) + "\t" + toString(count)); m->mothurOutEndLine(); }
- }
+ if (topdown) {
+ //think about running through twice...
+ for (int i = 0; i < numSeqs; i++) {
+
+ if (alignSeqs[i].active) { //this sequence has not been merged yet
+
+ string chunk = alignSeqs[i].seq.getName() + "\t" + toString(alignSeqs[i].numIdentical) + "\t" + toString(0) + "\t" + alignSeqs[i].seq.getAligned() + "\n";
+
+ //try to merge it with all smaller seqs
+ for (int j = i+1; j < numSeqs; j++) {
+
+ if (m->control_pressed) { out.close(); return 0; }
+
+ if (alignSeqs[j].active) { //this sequence has not been merged yet
+ //are you within "diff" bases
+ int mismatch = calcMisMatches(alignSeqs[i].seq.getAligned(), alignSeqs[j].seq.getAligned());
+
+ if (mismatch <= diffs) {
+ //merge
+ alignSeqs[i].names += ',' + alignSeqs[j].names;
+ alignSeqs[i].numIdentical += alignSeqs[j].numIdentical;
+
+ chunk += alignSeqs[j].seq.getName() + "\t" + toString(alignSeqs[j].numIdentical) + "\t" + toString(mismatch) + "\t" + alignSeqs[j].seq.getAligned() + "\n";
+
+ alignSeqs[j].active = 0;
+ alignSeqs[j].numIdentical = 0;
+ count++;
+ }
+ }//end if j active
+ }//end for loop j
+
+ //remove from active list
+ alignSeqs[i].active = 0;
+
+ out << "ideal_seq_" << (i+1) << '\t' << alignSeqs[i].numIdentical << endl << chunk << endl;;
+
+ }//end if active i
+ if(i % 100 == 0) { m->mothurOutJustToScreen(toString(i) + "\t" + toString(numSeqs - count) + "\t" + toString(count)+"\n"); }
+ }
+ }else {
+ map<int, string> mapFile;
+ map<int, int> originalCount;
+ map<int, int>::iterator itCount;
+ for (int i = 0; i < numSeqs; i++) { mapFile[i] = ""; originalCount[i] = alignSeqs[i].numIdentical; }
+
+ //think about running through twice...
+ for (int i = 0; i < numSeqs; i++) {
+
+ //try to merge it into larger seqs
+ for (int j = i+1; j < numSeqs; j++) {
+
+ if (m->control_pressed) { out.close(); return 0; }
+
+ if (originalCount[j] > originalCount[i]) { //this sequence is more abundant than I am
+ //are you within "diff" bases
+ int mismatch = calcMisMatches(alignSeqs[i].seq.getAligned(), alignSeqs[j].seq.getAligned());
+
+ if (mismatch <= diffs) {
+ //merge
+ alignSeqs[j].names += ',' + alignSeqs[i].names;
+ alignSeqs[j].numIdentical += alignSeqs[i].numIdentical;
+
+ mapFile[j] = alignSeqs[i].seq.getName() + "\t" + toString(alignSeqs[i].numIdentical) + "\t" + toString(mismatch) + "\t" + alignSeqs[i].seq.getAligned() + "\n" + mapFile[i];
+ alignSeqs[i].numIdentical = 0;
+ originalCount.erase(i);
+ mapFile[i] = "";
+ count++;
+ j+=numSeqs; //exit search, we merged this one in.
+ }
+ }//end abundance check
+ }//end for loop j
+
+ if(i % 100 == 0) { m->mothurOutJustToScreen(toString(i) + "\t" + toString(numSeqs - count) + "\t" + toString(count)+"\n"); }
+ }
+
+ for (int i = 0; i < numSeqs; i++) {
+ if (alignSeqs[i].numIdentical != 0) {
+ out << "ideal_seq_" << (i+1) << '\t' << alignSeqs[i].numIdentical << endl << alignSeqs[i].seq.getName() + "\t" + toString(alignSeqs[i].numIdentical) + "\t" + toString(0) + "\t" + alignSeqs[i].seq.getAligned() + "\n" << mapFile[i] << endl;
+ }
+ }
+
+ }
out.close();
if(numSeqs % 100 != 0) { m->mothurOut(toString(numSeqs) + "\t" + toString(numSeqs - count) + "\t" + toString(count)); m->mothurOutEndLine(); }
while (!in.eof()) {
in >> firstCol >> secondCol; m->gobble(in);
+
+ m->checkName(firstCol);
+ m->checkName(secondCol);
+ int size = m->getNumNames(secondCol);
+
names[firstCol] = secondCol;
- int size = 1;
-
- for(int i=0;i<secondCol.size();i++){
- if(secondCol[i] == ','){ size++; }
- }
- sizes[firstCol] = size;
+ sizes[firstCol] = size;
}
in.close();
}