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);
}
int VcfSplit::execute(int argc, char **argv)
{
String refFile = "";
String inputVcf = "";
String outputVcfBase = "";
String refName = "";
bool uncompress = false;
bool params = false;
bool noeof = false;
// Read in the parameters.
ParameterList inputParameters;
BEGIN_LONG_PARAMETERS(longParameterList)
LONG_PARAMETER_GROUP("Required Parameters")
LONG_STRINGPARAMETER("in", &inputVcf)
LONG_STRINGPARAMETER("obase", &outputVcfBase)
LONG_PARAMETER_GROUP("Optional Parameters")
LONG_PARAMETER("uncompress", &uncompress)
LONG_STRINGPARAMETER("refName", &refName)
LONG_PARAMETER("noeof", &noeof)
LONG_PARAMETER("params", ¶ms)
LONG_PHONEHOME(VERSION)
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(outputVcfBase == "")
{
usage();
inputParameters.Status();
std::cerr << "Missing \"--obase\", a required parameter.\n\n";
return(-1);
}
outputVcfBase += ".";
if(params)
{
inputParameters.Status();
}
// If no eof block is required for a bgzf file, set the bgzf file type to
// not look for it.
if(noeof)
{
// Set that the eof block is not required.
BgzfFileType::setRequireEofBlock(false);
}
VcfFileReader inFile;
std::map<std::string, VcfFileWriter*> outFiles;
VcfHeader header;
// Open the file.
inFile.open(inputVcf, header);
if(refName != "")
{
inFile.setReadSection(refName.c_str());
}
VcfRecord record;
int numRecords = 0;
std::string prevChr = "";
std::string chr = "";
VcfFileWriter* outFilePtr = 0;
std::string outName = "";
while(inFile.readRecord(record))
{
++numRecords;
chr = record.getChromStr();
if((outFilePtr == 0) || (chr != prevChr))
{
outFilePtr = outFiles[chr];
if(outFilePtr == 0)
{
outFilePtr = new VcfFileWriter();
outFiles[chr] = outFilePtr;
outName = outputVcfBase.c_str();
if(chr.substr(0,3) != "chr")
{
outName += "chr";
}
outName += chr + ".vcf";
// chr not in outFile list.
if(uncompress)
{
outFilePtr->open(outName.c_str(), header, InputFile::DEFAULT);
}
else
{
outName += ".gz";
outFilePtr->open(outName.c_str(), header);
}
}
}
outFilePtr->writeRecord(record);
}
inFile.close();
for (std::map<std::string,VcfFileWriter*>::iterator it = outFiles.begin();
it != outFiles.end(); ++it)
{
if(it->second != 0)
{
it->second->close();
it->second = 0;
}
}
std::cerr << "NumRecords: " << numRecords << "\n";
return(0);
}
int VcfConvert::execute(int argc, char **argv)
{
String refFile = "";
String inputVcf = "";
String outputVcf = "";
String refName = "";
bool uncompress = false;
bool params = false;
bool noeof = 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_STRINGPARAMETER("refName", &refName)
LONG_PARAMETER("noeof", &noeof)
LONG_PARAMETER("params", ¶ms)
LONG_PHONEHOME(VERSION)
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();
}
// If no eof block is required for a bgzf file, set the bgzf file type to
// not look for it.
if(noeof)
{
// Set that the eof block is not required.
BgzfFileType::setRequireEofBlock(false);
}
VcfFileReader inFile;
VcfFileWriter outFile;
VcfHeader header;
// Open the file.
inFile.open(inputVcf, header);
if(refName != "")
{
inFile.setReadSection(refName.c_str());
}
if(uncompress)
{
outFile.open(outputVcf, header, InputFile::DEFAULT);
}
else
{
outFile.open(outputVcf, header);
}
VcfRecord record;
int numRecords = 0;
while(inFile.readRecord(record))
{
++numRecords;
outFile.writeRecord(record);
}
inFile.close();
std::cerr << "NumRecords: " << numRecords << "\n";
return(0);
}
int VcfConsensus::execute(int argc, char ** argv)
{
String vcfName1;
String vcfName2;
String vcfName3;
String outputFileName;
bool uncompress = false;
bool params = false;
// Read in the parameters.
ParameterList inputParameters;
BEGIN_LONG_PARAMETERS(longParameterList)
LONG_PARAMETER_GROUP("Required Parameters")
LONG_STRINGPARAMETER("in1", &vcfName1)
LONG_STRINGPARAMETER("in2", &vcfName2)
LONG_STRINGPARAMETER("in3", &vcfName3)
LONG_STRINGPARAMETER("out", &outputFileName)
LONG_PARAMETER_GROUP("Optional Parameters")
LONG_PARAMETER("uncompress", &uncompress)
LONG_PARAMETER("params", ¶ms)
LONG_PHONEHOME(VERSION)
END_LONG_PARAMETERS();
inputParameters.Add(new LongParameters ("Input Parameters",
longParameterList));
inputParameters.Read(argc-1, &(argv[1]));
std::string gtField = "GT";
VcfFileReader vcf1;
VcfFileReader vcf2;
VcfFileReader vcf3;
VcfHeader header1;
VcfHeader header2;
VcfHeader header3;
VcfRecord record1;
VcfRecord record2;
VcfRecord record3;
VcfRecordGenotype* genotypeInfoPtr1 = NULL;
VcfRecordGenotype* genotypeInfoPtr2 = NULL;
VcfRecordGenotype* genotypeInfoPtr3 = NULL;
unsigned int numMissing2 = 0;
unsigned int numMissing3 = 0;
unsigned int numMismatchRefAlt = 0;
unsigned int numMissingGT1 = 0;
const unsigned int myMaxErrors = 4;
// Check that the required parameters were set.
if(vcfName1 == "")
{
usage();
inputParameters.Status();
std::cerr << "Missing \"--in1\", a required parameter.\n\n";
return(-1);
}
if(vcfName2 == "")
{
usage();
inputParameters.Status();
std::cerr << "Missing \"--in2\", a required parameter.\n\n";
return(-1);
}
if(vcfName3 == "")
{
usage();
inputParameters.Status();
std::cerr << "Missing \"--in3\", a required parameter.\n\n";
return(-1);
}
if(outputFileName == "")
{
usage();
inputParameters.Status();
std::cerr << "Missing \"--out\", a required parameter.\n\n";
return(-1);
}
if(params)
{
inputParameters.Status();
}
// Open the files.
vcf1.open(vcfName1, header1);
vcf2.open(vcfName2, header2);
vcf3.open(vcfName3, header3);
// Setup the sample name maps.
int numSamples = header1.getNumSamples();
std::vector<int> sample2Indices;
std::vector<int> sample3Indices;
std::vector<int> removeIndices;
int numSamplesSkipped1 = 0;
int numSamplesSkipped2 = 0;
int numSamplesSkipped3 = 0;
for(int i = 0; i < numSamples; i++)
{
int sm2Index = header2.getSampleIndex(header1.getSampleName(i));
int sm3Index = header3.getSampleIndex(header1.getSampleName(i));
// Look for this sample name in vcf2.
if((sm2Index != -1) && (sm3Index != -1))
{
sample2Indices.push_back(sm2Index);
sample3Indices.push_back(sm3Index);
}
else
{
// Sample not found in all three vcfs.
removeIndices.push_back(i);
++numSamplesSkipped1;
}
}
// Remove samples not found in all 3 vcfs from header1.
// Remove them in reverse order so they are removed from the end of the header first.
VcfSubsetSamples subset1;
subset1.init(header1, true);
for(int i = (removeIndices.size() - 1); i >= 0; i--)
{
subset1.addExcludeSample(header1.getSampleName(removeIndices[i]));
header1.removeSample(removeIndices[i]);
}
// Set numSamples to the new number of samples in header1.
numSamples = header1.getNumSamples();
// Calculate the number of samples skipped for files 2 & 3.
numSamplesSkipped2 = header2.getNumSamples() - sample2Indices.size();
numSamplesSkipped3 = header3.getNumSamples() - sample3Indices.size();
if(numSamplesSkipped1 > 0)
{
std::cerr << "Skipping " << numSamplesSkipped1 << " samples from --in1\n";
}
if(numSamplesSkipped2 > 0)
{
std::cerr << "Skipping " << numSamplesSkipped2 << " samples from --in2\n";
}
if(numSamplesSkipped3 > 0)
{
std::cerr << "Skipping " << numSamplesSkipped3 << " samples from --in3\n";
}
VcfFileWriter outputVcf;
// Open and write the header
if(uncompress)
{
outputVcf.open(outputFileName, header1, InputFile::DEFAULT);
}
else
{
outputVcf.open(outputFileName, header1);
}
const char* chrom1 = NULL;
int pos1 = UNSET_POS;
// Read the first record from vcf2 & vcf3.
vcf2.readRecord(record2);
vcf3.readRecord(record3);
bool newChrom = true;
static std::string prevChrom = "";
uint64_t numAllMatch = 0;
uint64_t num1Match2Only = 0;
uint64_t num1Match3Only = 0;
uint64_t num2Match3Only = 0;
uint64_t numNoMatches = 0;
uint64_t numAllMatch00 = 0;
uint64_t num1Match2Only00 = 0;
uint64_t num1Match3Only00 = 0;
uint64_t num2Match3Only00 = 0;
uint64_t numAllMatch01 = 0;
uint64_t num1Match2Only01 = 0;
uint64_t num1Match3Only01 = 0;
uint64_t num2Match3Only01 = 0;
uint64_t numAllMatch11 = 0;
uint64_t num1Match2Only11 = 0;
uint64_t num1Match3Only11 = 0;
uint64_t num2Match3Only11 = 0;
// Loop through vcf1.
while(vcf1.readRecord(record1, &subset1))
{
chrom1 = record1.getChromStr();
pos1 = record1.get1BasedPosition();
if(strcmp(chrom1, prevChrom.c_str()) == 0)
{
newChrom = false;
}
else
{
prevChrom = chrom1;
newChrom = true;
}
bool found = true;
if(!findPos(newChrom, chrom1, pos1, record2, vcf2))
{
// Failed to find the position, continue to the next position
if(++numMissing2 <= myMaxErrors)
{
std::cerr << "Failed to find " << chrom1 << ":" << pos1
<< " in " << vcfName2 << ", so skipping this pos\n";
}
found = false;
}
if(!findPos(newChrom, chrom1, pos1, record3, vcf3))
{
// Failed to find the position, continue to the next position
if(++numMissing3 <= myMaxErrors)
{
std::cerr << "Failed to find " << chrom1 << ":" << pos1
<< " in " << vcfName3 << ", so skipping this pos\n";
}
found = false;
}
if(found == false)
{
continue;
}
// Found the position in all files.
// Validate that the reference & alternate alleles are the same.
const char* ref1 = record1.getRefStr();
const char* alt1 = record1.getAltStr();
if((strcmp(ref1, record2.getRefStr()) != 0) ||
(strcmp(ref1, record3.getRefStr()) != 0) ||
(strcmp(alt1, record2.getAltStr()) != 0) ||
(strcmp(alt1, record3.getAltStr()) != 0))
{
if(++numMismatchRefAlt <= myMaxErrors)
{
std::cerr << "Mismatching ref/alt found at " << chrom1 << ":" << pos1 << ", so skipping this pos\n";
}
continue;
}
// Get the genotype information for each.
genotypeInfoPtr1 = &(record1.getGenotypeInfo());
genotypeInfoPtr2 = &(record2.getGenotypeInfo());
genotypeInfoPtr3 = &(record3.getGenotypeInfo());
// Loop through all the samples in vcf1.
// Get the Genotype Information.
for(int i = 0; i < numSamples; i++)
{
const std::string* genotypeVal1 = genotypeInfoPtr1->getString(gtField, i);
const std::string* genotypeVal2 = genotypeInfoPtr2->getString(gtField, sample2Indices[i]);
const std::string* genotypeVal3 = genotypeInfoPtr3->getString(gtField, sample3Indices[i]);
// Need to make sure the field was found.
if(genotypeVal1 == NULL)
{
// GT not found in the first record, so just continue.
if(++numMissingGT1 <= myMaxErrors)
{
std::cerr << "Missing GT for " << header1.getSampleName(i) << " in " << vcfName1 << "\n";
}
continue;
}
if(isSame(genotypeVal1, genotypeVal2))
{
// genotypeVal1 is majority, so make no change.
if(isSame(genotypeVal1, genotypeVal3))
{
++numAllMatch;
if(*genotypeVal1 == "0/0")
{
++numAllMatch00;
}
else if((*genotypeVal1 == "0/1") ||
(*genotypeVal1 == "1/0"))
{
++numAllMatch01;
}
if(*genotypeVal1 == "1/1")
{
++numAllMatch11;
}
}
else
{
++num1Match2Only;
if(*genotypeVal1 == "0/0")
{
++num1Match2Only00;
}
else if((*genotypeVal1 == "0/1") ||
(*genotypeVal1 == "1/0"))
{
++num1Match2Only01;
}
if(*genotypeVal1 == "1/1")
{
++num1Match2Only11;
}
}
}
else if(isSame(genotypeVal1, genotypeVal3))
{
// genotypeVal1 is majority, so make no change.
++num1Match3Only;
if(*genotypeVal1 == "0/0")
{
++num1Match3Only00;
}
else if((*genotypeVal1 == "0/1") ||
(*genotypeVal1 == "1/0"))
{
++num1Match3Only01;
}
if(*genotypeVal1 == "1/1")
{
++num1Match3Only11;
}
}
else if(isSame(genotypeVal2, genotypeVal3))
{
// genotypeVal2 is majority, so change genotypeVal1.
genotypeInfoPtr1->setString(gtField, i, *genotypeVal2);
++num2Match3Only;
if(*genotypeVal2 == "0/0")
{
++num2Match3Only00;
}
else if((*genotypeVal2 == "0/1") ||
(*genotypeVal2 == "1/0"))
{
++num2Match3Only01;
}
if(*genotypeVal2 == "1/1")
{
++num2Match3Only11;
}
}
else
{
// None are the same so set to "./."
genotypeInfoPtr1->setString(gtField, i, "./.");
++numNoMatches;
}
} // loop back to vcf1 samples.
// Write this record.
outputVcf.writeRecord(record1);
} // loop back to next vcf1 record.
std::cerr << "\n";
if(numMissing2 > myMaxErrors)
{
std::cerr << "Suppressed "
<< numMissing2 - myMaxErrors
<< " errors about skipped positions because they were not in "
<< vcfName2
<< "\n";
}
if(numMissing3 > myMaxErrors)
{
std::cerr << "Suppressed "
<< numMissing3 - myMaxErrors
<< " errors about skipped positions because they were not in "
<< vcfName3
<< "\n";
}
if(numMismatchRefAlt > myMaxErrors)
{
std::cerr << "Suppressed "
<< numMismatchRefAlt - myMaxErrors
<< " errors about mismatched ref/alt positions\n";
}
if(numMissingGT1 > myMaxErrors)
{
std::cerr << "Suppressed "
<< numMissingGT1 - myMaxErrors
<< " errors about missing GT for "
<< vcfName1
<< "\n";
}
std::cerr << "\n";
// Output the stats.
std::cerr << "File1 = " << vcfName1 << std::endl;
std::cerr << "File2 = " << vcfName2 << std::endl;
std::cerr << "File3 = " << vcfName3 << std::endl;
std::cerr << "\nType\tTotal\t0/0\t0/1|1/0\t1/1\n";
std::cerr << "AllMatched"
<< "\t" << numAllMatch
<< "\t" << numAllMatch00
<< "\t" << numAllMatch01
<< "\t" << numAllMatch11 << std::endl;
std::cerr << "1matched2"
<< "\t" << num1Match2Only
<< "\t" << num1Match2Only00
<< "\t" << num1Match2Only01
<< "\t" << num1Match2Only11 << std::endl;
std::cerr << "1matched3"
<< "\t" << num1Match3Only
<< "\t" << num1Match3Only00
<< "\t" << num1Match3Only01
<< "\t" << num1Match3Only11 << std::endl;
std::cerr << "2matched3"
<< "\t" << num2Match3Only
<< "\t" << num2Match3Only00
<< "\t" << num2Match3Only01
<< "\t" << num2Match3Only11 << std::endl;
std::cerr << "NoneMatched\t" << numNoMatches << std::endl;
return(0);
}
