uint loadGTF(SjdbClass &sjdbLoci, Parameters *P) {//load gtf file, add junctions to P->sjdb
//returns number of added junctions
if (P->sjdbOverhang>0 && P->sjdbGTFfile!="-") {
ifstream sjdbStreamIn ( P->sjdbGTFfile.c_str() );
if (sjdbStreamIn.fail()) {
ostringstream errOut;
errOut << "FATAL error, could not open file sjdbGTFfile=" << P->sjdbGTFfile <<"\n";
exitWithError(errOut.str(),std::cerr, P->inOut->logMain, EXIT_CODE_INPUT_FILES, *P);
};
std::map <string,uint> transcriptIDnumber;
uint exonN=0;
while (sjdbStreamIn.good()) {//count the number of exons
string chr1,ddd2,featureType;
sjdbStreamIn >> chr1 >> ddd2 >> featureType;
if (chr1.substr(0,1)!="#" && featureType==P->sjdbGTFfeatureExon) {
exonN++;
};
sjdbStreamIn.ignore(1000000000,'\n'); //ignore the rest of the line
};
uint* exonLoci=new uint [exonN*GTF_exonLoci_size];
char* transcriptStrand = new char [exonN];
vector <string> transcriptID;
exonN=0;//re-calculate
sjdbStreamIn.clear();
sjdbStreamIn.seekg(0,ios::beg);
while (sjdbStreamIn.good()) {
string oneLine,chr1,ddd2,featureType;
getline(sjdbStreamIn,oneLine);
istringstream oneLineStream (oneLine);
oneLineStream >> chr1 >> ddd2 >> featureType;
if (chr1.substr(0,1)!="#" && featureType==P->sjdbGTFfeatureExon) {//exonic line, process
if (P->sjdbGTFchrPrefix!="-") chr1=P->sjdbGTFchrPrefix + chr1;
if (P->chrNameIndex.count(chr1)==0) {//chr not in Genome
P->inOut->logMain << "WARNING: while processing sjdbGTFfile=" << P->sjdbGTFfile <<": chromosome '"<<chr1<<"' not found in Genome fasta files for line:\n";
P->inOut->logMain << oneLine <<"\n"<<flush;
continue; //do not process exons/transcripts on missing chromosomes
};
uint ex1,ex2;
char str1;
oneLineStream >> ex1 >> ex2 >> ddd2 >> str1 >> ddd2; //read all fields except the last
string oneLine1;
getline(oneLineStream, oneLine1);//get the last field
replace(oneLine1.begin(),oneLine1.end(),';',' ');//to separate attributes
replace(oneLine1.begin(),oneLine1.end(),'=',' ');//for GFF3 processing
oneLineStream.str(oneLine1);
oneLineStream.clear();
string trID(""), attr1("");
while (oneLineStream.good()) {
oneLineStream >> attr1;
if (attr1==P->sjdbGTFtagExonParentTranscript) {
oneLineStream >> trID;
trID.erase(remove(trID.begin(),trID.end(),'"'),trID.end());
trID.erase(remove(trID.begin(),trID.end(),';'),trID.end());
// cout <<trID<<endl;
};
};
if (trID=="") {//no transcript ID
P->inOut->logMain << "WARNING: while processing sjdbGTFfile=" << P->sjdbGTFfile <<": no transcript_id for exon feature for line:\n";
P->inOut->logMain << oneLine <<"\n"<<flush;
} else {
transcriptIDnumber.insert(std::pair <string,uint> (trID,(uint) transcriptIDnumber.size()));//insert new element if necessary with a new numeric value
if (transcriptID.size() < transcriptIDnumber.size()) transcriptID.push_back(trID);
if (str1=='+') {
transcriptStrand[transcriptIDnumber[trID]]=1;
} else if (str1=='-') {
transcriptStrand[transcriptIDnumber[trID]]=2;
} else {
transcriptStrand[transcriptIDnumber[trID]]=0;
};
};
exonLoci[GTF_exonTrID(exonN)]=transcriptIDnumber[trID];
exonLoci[GTF_exonStart(exonN)]=ex1+P->chrStart[P->chrNameIndex[chr1]]-1;
exonLoci[GTF_exonEnd(exonN)]=ex2+P->chrStart[P->chrNameIndex[chr1]]-1;
++exonN;
};//if (chr1.substr(0,1)!="#" && featureType=="exon")
};//
uint loadGTF(SjdbClass &sjdbLoci, Parameters *P, string dirOut) {//load gtf file, add junctions to P->sjdb
//returns number of added junctions
if (P->sjdbOverhang>0 && P->sjdbGTFfile!="-") {
time_t rawTime;
time(&rawTime);
P->inOut->logMain << timeMonthDayTime(rawTime) <<" ..... Processing annotations GTF\n" <<flush;
*P->inOut->logStdOut << timeMonthDayTime(rawTime) <<" ..... Processing annotations GTF\n" <<flush;
ifstream sjdbStreamIn ( P->sjdbGTFfile.c_str() );
if (sjdbStreamIn.fail()) {
ostringstream errOut;
errOut << "FATAL error, could not open file sjdbGTFfile=" << P->sjdbGTFfile <<"\n";
exitWithError(errOut.str(),std::cerr, P->inOut->logMain, EXIT_CODE_INPUT_FILES, *P);
};
if (P->chrNameIndex.size()==0)
{
for (uint ii=0;ii<P->nChrReal;ii++) {
P->chrNameIndex[P->chrName[ii]]=ii;
};
};
std::map <string,uint> transcriptIDnumber, geneIDnumber;
uint exonN=0;
while (sjdbStreamIn.good()) {//count the number of exons
string chr1,ddd2,featureType;
sjdbStreamIn >> chr1 >> ddd2 >> featureType;
if (chr1.substr(0,1)!="#" && featureType==P->sjdbGTFfeatureExon) {
exonN++;
};
sjdbStreamIn.ignore(1000000000,'\n'); //ignore the rest of the line
};
if (exonN==0)
{
P->inOut->logMain << "WARNING: found no exons in sjdbGTFfile=" << P->sjdbGTFfile <<endl;
return 0;
};
uint* exonLoci=new uint [exonN*GTF_exonLoci_size];
char* transcriptStrand = new char [exonN];
vector <string> transcriptID, geneID;
exonN=0;//re-calculate
sjdbStreamIn.clear();
sjdbStreamIn.seekg(0,ios::beg);
while (sjdbStreamIn.good()) {
string oneLine,chr1,ddd2,featureType;
getline(sjdbStreamIn,oneLine);
istringstream oneLineStream (oneLine);
oneLineStream >> chr1 >> ddd2 >> featureType;
if (chr1.substr(0,1)!="#" && featureType==P->sjdbGTFfeatureExon) {//exonic line, process
if (P->sjdbGTFchrPrefix!="-") chr1=P->sjdbGTFchrPrefix + chr1;
if (P->chrNameIndex.count(chr1)==0) {//chr not in Genome
P->inOut->logMain << "WARNING: while processing sjdbGTFfile=" << P->sjdbGTFfile <<": chromosome '"<<chr1<<"' not found in Genome fasta files for line:\n";
P->inOut->logMain << oneLine <<"\n"<<flush;
continue; //do not process exons/transcripts on missing chromosomes
};
uint ex1,ex2;
char str1;
oneLineStream >> ex1 >> ex2 >> ddd2 >> str1 >> ddd2; //read all fields except the last
string oneLine1;
getline(oneLineStream, oneLine1);//get the last field
replace(oneLine1.begin(),oneLine1.end(),';',' ');//to separate attributes
replace(oneLine1.begin(),oneLine1.end(),'=',' ');//for GFF3 processing
oneLineStream.str(oneLine1);
oneLineStream.clear();
string trID(""), gID(""), attr1("");
while (oneLineStream.good()) {
oneLineStream >> attr1;
if (attr1==P->sjdbGTFtagExonParentTranscript) {
oneLineStream >> trID;
trID.erase(remove(trID.begin(),trID.end(),'"'),trID.end());
trID.erase(remove(trID.begin(),trID.end(),';'),trID.end());
} else if (attr1==P->sjdbGTFtagExonParentGene) {
oneLineStream >> gID;
gID.erase(remove(gID.begin(),gID.end(),'"'),gID.end());
gID.erase(remove(gID.begin(),gID.end(),';'),gID.end());
};
};
