int VcfCleaner::execute(int argc, char **argv)
{
String refFile = "";
String inputVcf = "";
String outputVcf = "";
bool uncompress = false;
bool params = false;
// Read in the parameters.
ParameterList inputParameters;
BEGIN_LONG_PARAMETERS(longParameterList)
LONG_PARAMETER_GROUP("Required Parameters")
LONG_STRINGPARAMETER("in", &inputVcf)
LONG_STRINGPARAMETER("out", &outputVcf)
LONG_PARAMETER_GROUP("Optional Parameters")
LONG_PARAMETER("uncompress", &uncompress)
LONG_PARAMETER("params", ¶ms)
END_LONG_PARAMETERS();
inputParameters.Add(new LongParameters ("Input Parameters",
longParameterList));
inputParameters.Read(argc-1, &(argv[1]));
// Check that all files were specified.
if(inputVcf == "")
{
usage();
inputParameters.Status();
std::cerr << "Missing \"--in\", a required parameter.\n\n";
return(-1);
}
if(outputVcf == "")
{
usage();
inputParameters.Status();
std::cerr << "Missing \"--out\", a required parameter.\n\n";
return(-1);
}
if(params)
{
inputParameters.Status();
}
VcfFileReader inFile;
VcfFileWriter outFile;
VcfHeader header;
VcfRecord record;
// Open the file.
inFile.open(inputVcf, header);
if(uncompress)
{
outFile.open(outputVcf, header, InputFile::DEFAULT);
}
else
{
outFile.open(outputVcf, header);
}
int numReadRecords = 0;
int numWrittenRecords = 0;
int returnVal = 0;
// Set to only store/write the GT field.
VcfRecordGenotype::addStoreField("GT");
while(inFile.readRecord(record))
{
++numReadRecords;
// Check if any samples are missing GT or if any are not phased.
if(!record.hasAllGenotypeAlleles() || !record.allPhased())
{
// Missing a GT or not phased, so continue without writing.
continue;
}
// Clear the INFO field.
record.getInfo().clear();
// Write the record.
if(!outFile.writeRecord(record))
{
// Write error.
std::cerr << "Failed writing a vcf record.\n";
returnVal = -1;
}
++numWrittenRecords;
}
inFile.close();
outFile.close();
std::cerr << "NumReadRecords: " << numReadRecords
<< "; NumWrittenRecords: " << numWrittenRecords << "\n";
return(returnVal);
}
bool VcfFileReader::readRecord(VcfRecord& record, VcfSubsetSamples* subset)
{
myStatus = StatGenStatus::SUCCESS;
// Subset the read if there are subsets specified.
VcfSubsetSamples* subsetPtr = subset;
if((subsetPtr == NULL) && myUseSubset)
{
subsetPtr = &mySampleSubset;
}
// Check to see if a new region has been set. If so, setup for that region.
bool searchChrom = false;
if(myNewSection)
{
if(myVcfIndex != NULL)
{
// Have an index file so use
if(!processNewSection())
{
// processNewSection sets the status appropriately on failure.
return(false);
}
}
else if(myTotalRead == 0)
{
// ReadSection without an index only works if no records
// have been read.
searchChrom = true;
myNewSection = false;
}
else
{
myNewSection = false;
myStatus.setStatus(StatGenStatus::FAIL_ORDER,
"Cannot set read section with no index after reading records");
return(false);
}
}
// Keep looping until a desired record is found.
bool recordFound = false;
while(!recordFound)
{
if(!record.read(myFilePtr, mySiteOnly, myRecordDiscardRules, subsetPtr))
{
myStatus = record.getStatus();
myTotalRead += myRecordDiscardRules.getNumDiscarded();
myNumRecords += myRecordDiscardRules.getNumDiscarded();
myRecordDiscardRules.clearNumDiscarded();
return(false);
}
++myTotalRead;
myTotalRead += myRecordDiscardRules.getNumDiscarded();
// Check to see if the record is in the section.
// First check the chromosome.
if(!mySectionChrom.empty() && (mySectionChrom != record.getChromStr()))
{
if(searchChrom)
{
// Still searching for the chromosome, so continue
// to the next record.
continue;
}
// Record is not within the correct chromosome, so return failure.
myStatus = StatGenStatus::NO_MORE_RECS;
return(false);
}
searchChrom = false;
// Check if the record is after the section end if applicable.
if((mySection1BasedEndPos != -1) &&
(record.get1BasedPosition() >= mySection1BasedEndPos))
{
myStatus = StatGenStatus::NO_MORE_RECS;
return(false);
}
// Check if the record is prior to the section start if applicable.
// Determinine the VCF record end position.
// If we are not requiring overlap, then we only need to check
// the start position, but if overlap is required, then it needs
// to incrment the start by the length-1.
int numIncBases = 0;
if(mySectionOverlap)
{
// The VCF record end position is the start position + length of the
// reference string - 1.
numIncBases = record.getNumRefBases() - 1;
}
if((mySection1BasedStartPos != -1) &&
((record.get1BasedPosition() + numIncBases)
< mySection1BasedStartPos))
{
// This record is prior to the section, so keep reading.
continue;
}
++myNumRecords;
myNumRecords += myRecordDiscardRules.getNumDiscarded();
myRecordDiscardRules.clearNumDiscarded();
// Record successfully read, so check to see if it is discarded.
if((myDiscardRules & DISCARD_NON_PHASED) && !record.allPhased())
{
// Not all samples are phased, so discard this record.
continue;
}
if((myDiscardRules & DISCARD_MISSING_GT) &&
!record.hasAllGenotypeAlleles())
{
// discard missing GTs and this record had missing alleles,
// so keep reading.
continue;
}
if((myDiscardRules & DISCARD_FILTERED) &&
!(record.getFilter().passedAllFilters()))
{
// Record was filtered, so discard it.
continue;
}
if((myDiscardRules & DISCARD_MULTIPLE_ALTS) &&
(record.getNumAlts() > 1))
{
// Record had multiple alternates, so discard.
continue;
}
// Check allele counts for discarding.
if(myMinAltAlleleCount != UNSET_MIN_ALT_ALLELE_COUNT)
{
// Count the number of alternates.
int32_t altCount = 0;
for(int sampleNum = 0; sampleNum < record.getNumSamples();
sampleNum++)
{
if((myAltAlleleCountSubset != NULL) &&
!(myAltAlleleCountSubset->keep(sampleNum)))
{
// Skip this sample.
continue;
}
for(int gtNum = 0; gtNum < record.getNumGTs(sampleNum); gtNum++)
{
if(record.getGT(sampleNum, gtNum) > 0)
{
// Alternate, so increment the count.
++altCount;
}
}
}
if(altCount < myMinAltAlleleCount)
{
// Not enough alternates so continue to the next sample.
continue;
}
}
// Check to see if the minimum alternate allele count is met.
if(myMinMinorAlleleCount != UNSET_MIN_MINOR_ALLELE_COUNT)
{
// Get the number of possible alternates.
unsigned int numAlts = record.getNumAlts();
// Verify that each allele has the min count.
bool failMinorAlleleCount = false;
for(unsigned int i = 0; i <= numAlts; i++)
{
if(record.getAlleleCount(i, myMinorAlleleCountSubset)
< myMinMinorAlleleCount)
{
// Not enough of one gt, so not ok.
failMinorAlleleCount = true;
break;
}
}
if(failMinorAlleleCount)
{
// not enough alleles, so continue to the next record.
continue;
}
}
// Record was not discarded.
recordFound = true;
}
// Increment the number of kept records.
++myNumKeptRecords;
return(true);
}