void testAddHeaderAndTagToFile(const char* inputName, const char* outputName)
{
SamFile inSam, outSam;
assert(inSam.OpenForRead(inputName));
assert(outSam.OpenForWrite(outputName));
// Read the SAM Header.
SamFileHeader samHeader;
assert(inSam.ReadHeader(samHeader));
// Add a header line.
assert(samHeader.addHeaderLine("@RG\tID:myID\tSM:mySM") == false);
assert(samHeader.addHeaderLine("@RG\tID:myID3\tSM:mySM") == true);
// Write Header
assert(outSam.WriteHeader(samHeader));
SamRecord samRecord;
assert(inSam.ReadRecord(samHeader, samRecord));
// validateRead1(samRecord);
// Add two tags.
assert(samRecord.addIntTag("XA", 123));
assert(samRecord.addIntTag("XA", 456));
assert(samRecord.addTag("RR", 'Z', "myID1"));
assert(samRecord.addTag("RR", 'Z', "myID2"));
// Write as Sam.
assert(outSam.WriteRecord(samHeader, samRecord));
// TODO, add test to verify it was written correctly.
// Read a couple of records to make sure it properly can read them even
// if they are bigger than the original.
assert(inSam.ReadRecord(samHeader, samRecord));
assert(inSam.ReadRecord(samHeader, samRecord));
// Check the MD tag, which requires the reference.
GenomeSequence reference("testFiles/chr1_partial.fa");
assert(SamTags::isMDTagCorrect(samRecord, reference) == false);
String newMDTag;
SamTags::createMDTag(newMDTag, samRecord, reference);
assert(newMDTag == "2T1N0");
assert(SamTags::updateMDTag(samRecord, reference));
// Write as Sam.
assert(outSam.WriteRecord(samHeader, samRecord));
}
int main(int argc, char ** argv)
{
gpLogger = new Logger;
static struct option getopt_long_options[] =
{
// Input options
{ "fasta", required_argument, NULL, 'f'},
{ "in", required_argument, NULL, 'i'},
{ "out", required_argument, NULL, 'o'},
{ "verbose", no_argument, NULL, 'v'},
{ "log", required_argument, NULL, 'l'},
{ "clear", no_argument, NULL, 0},
{ "AS", required_argument, NULL, 0},
{ "UR", required_argument, NULL, 0},
{ "SP", required_argument, NULL, 0},
{ "HD", required_argument, NULL, 0},
{ "RG", required_argument, NULL, 0},
{ "PG", required_argument, NULL, 0},
{ "checkSQ", no_argument, NULL, 0},
{ NULL, 0, NULL, 0 },
};
int n_option_index = 0, c;
std::string sAS, sUR, sSP, sFasta, sInFile, sOutFile, sLogFile;
bool bClear, bCheckSQ, bVerbose;
std::vector<std::string> vsHDHeaders, vsRGHeaders, vsPGHeaders;
bCheckSQ = bVerbose = false;
bClear = true;
while ( (c = getopt_long(argc, argv, "vf:i:o:l:", getopt_long_options, &n_option_index)) != -1 ) {
// std::cout << getopt_long_options[n_option_index].name << "\t" << optarg << std::endl;
if ( c == 'f' ) {
sFasta = optarg;
}
else if ( c == 'i' ) {
sInFile = optarg;
}
else if ( c == 'o' ) {
sOutFile = optarg;
}
else if ( c == 'v' ) {
bVerbose = true;
}
else if ( c == 'l' ) {
sLogFile = optarg;
}
else if ( strcmp(getopt_long_options[n_option_index].name,"AS") == 0 ) {
sAS = optarg;
}
else if ( strcmp(getopt_long_options[n_option_index].name,"UR") == 0 ) {
sUR = optarg;
}
else if ( strcmp(getopt_long_options[n_option_index].name,"SP") == 0 ) {
sSP = optarg;
}
else if ( strcmp(getopt_long_options[n_option_index].name,"HD") == 0 ) {
vsHDHeaders.push_back(optarg);
}
else if ( strcmp(getopt_long_options[n_option_index].name,"RG") == 0 ) {
vsRGHeaders.push_back(optarg);
}
else if ( strcmp(getopt_long_options[n_option_index].name,"PG") == 0 ) {
vsPGHeaders.push_back(optarg);
}
else if ( strcmp(getopt_long_options[n_option_index].name,"checkSQ") == 0 ) {
bCheckSQ = true;
}
else {
std::cerr << "Error: Unrecognized option " << getopt_long_options[n_option_index].name << std::endl;
abort();
}
}
if ( optind < argc ) {
printUsage(std::cerr);
gpLogger->error("non-option argument %s exist ",argv[optind]);
}
if ( sInFile.empty() || sOutFile.empty() ) {
printUsage(std::cerr);
gpLogger->error("Input and output files are required");
}
if ( sLogFile.compare("__NONE__") == 0 ) {
sLogFile = (sOutFile + ".log");
}
gpLogger->open(sLogFile.c_str(), bVerbose);
if ( ( bCheckSQ ) && ( sFasta.empty() ) ) {
printUsage(std::cerr);
gpLogger->error("--checkSQ option must be used with --fasta option");
}
// check whether each header line starts with a correct tag
checkHeaderStarts(vsHDHeaders, "@HD\t");
checkHeaderStarts(vsRGHeaders, "@RG\t");
checkHeaderStarts(vsPGHeaders, "@PG\t");
gpLogger->write_log("Arguments in effect:");
gpLogger->write_log("\t--in [%s]",sInFile.c_str());
gpLogger->write_log("\t--out [%s]",sOutFile.c_str());
gpLogger->write_log("\t--log [%s]",sLogFile.c_str());
gpLogger->write_log("\t--fasta [%s]",sFasta.c_str());
gpLogger->write_log("\t--AS [%s]",sAS.c_str());
gpLogger->write_log("\t--UR [%s]",sUR.c_str());
gpLogger->write_log("\t--SP [%s]",sSP.c_str());
gpLogger->write_log("\t--checkSQ [%s]",bClear ? "ON" : "OFF" );
if ( vsHDHeaders.empty() ) {
gpLogger->write_log("\t--HD []");
}
else {
gpLogger->write_log("\t--HD [%s]",vsHDHeaders[0].c_str());
}
if ( vsRGHeaders.empty() ) {
gpLogger->write_log("\t--RG []");
}
else {
gpLogger->write_log("\t--RG [%s]",vsRGHeaders[0].c_str());
}
if ( vsPGHeaders.empty() ) {
gpLogger->write_log("\t--PG []");
}
else {
for(uint32_t i=0; i < vsPGHeaders.size(); ++i) {
gpLogger->write_log("\t--PG [%s]",vsPGHeaders[i].c_str());
}
}
if ( (vsHDHeaders.empty() ) && ( vsRGHeaders.empty() ) && ( vsPGHeaders.empty() ) && ( !bClear ) && ( sFasta.empty() ) ) {
gpLogger->warning("No option is in effect for modifying BAM files. The input and output files will be identical");
}
if ( ( vsHDHeaders.size() > 1 ) || ( vsRGHeaders.size() > 1 ) ) {
gpLogger->error("HD and RG headers cannot be multiple");
}
FastaFile fastaFile;
if ( ! sFasta.empty() ) {
if ( fastaFile.open(sFasta.c_str()) ) {
gpLogger->write_log("Reading the reference file %s",sFasta.c_str());
fastaFile.readThru();
fastaFile.close();
gpLogger->write_log("Finished reading the reference file %s",sFasta.c_str());
}
else {
gpLogger->error("Failed to open reference file %s",sFasta.c_str());
}
}
SamFile samIn;
SamFile samOut;
if ( ! samIn.OpenForRead(sInFile.c_str()) ) {
gpLogger->error("Cannot open BAM file %s for reading - %s",sInFile.c_str(), SamStatus::getStatusString(samIn.GetStatus()) );
}
if ( ! samOut.OpenForWrite(sOutFile.c_str()) ) {
gpLogger->error("Cannot open BAM file %s for writing - %s",sOutFile.c_str(), SamStatus::getStatusString(samOut.GetStatus()) );
}
SamFileHeader samHeader;
SamHeaderRecord* pSamHeaderRecord;
samIn.ReadHeader(samHeader);
// check the sanity of SQ file
// make sure the SN and LN matches, with the same order
if ( bCheckSQ ) {
unsigned int numSQ = 0;
while( (pSamHeaderRecord = samHeader.getNextHeaderRecord()) != NULL ) {
if ( pSamHeaderRecord->getType() == SamHeaderRecord::SQ ) {
++numSQ;
}
}
if ( numSQ != fastaFile.vsSequenceNames.size() ) {
gpLogger->error("# of @SQ tags are different from the original BAM and the reference file");
}
// iterator over all @SQ objects
for(unsigned int i=0; i < numSQ; ++i) {
pSamHeaderRecord = samHeader.getSQ(fastaFile.vsSequenceNames[i].c_str());
if ( fastaFile.vsSequenceNames[i].compare(pSamHeaderRecord->getTagValue("SN")) != 0 ) {
gpLogger->error("SequenceName is not identical between fasta and input BAM file");
}
else if ( static_cast<int>(fastaFile.vnSequenceLengths[i]) != atoi(pSamHeaderRecord->getTagValue("LN")) ) {
gpLogger->error("SequenceLength is not identical between fasta and input BAM file");
}
else {
if ( !sAS.empty() )
samHeader.setSQTag("AS",sAS.c_str(),fastaFile.vsSequenceNames[i].c_str());
samHeader.setSQTag("M5",fastaFile.vsMD5sums[i].c_str(),fastaFile.vsSequenceNames[i].c_str());
if ( !sUR.empty() )
samHeader.setSQTag("UR",sUR.c_str(),fastaFile.vsSequenceNames[i].c_str());
if ( !sSP.empty() )
samHeader.setSQTag("SP",sSP.c_str(),fastaFile.vsSequenceNames[i].c_str());
}
}
gpLogger->write_log("Finished checking the consistency of SQ tags");
}
else {
gpLogger->write_log("Skipped checking the consistency of SQ tags");
}
// go over the headers again,
// assuming order of HD, SQ, RG, PG, and put proper tags at the end of the original tags
gpLogger->write_log("Creating the header of new output file");
//SamFileHeader outHeader;
samHeader.resetHeaderRecordIter();
for(unsigned int i=0; i < vsHDHeaders.size(); ++i) {
samHeader.addHeaderLine(vsHDHeaders[i].c_str());
}
/*
for(int i=0; i < fastaFile.vsSequenceNames.size(); ++i) {
std::string s("@SQ\tSN:");
char buf[1024];
s += fastaFile.vsSequenceNames[i];
sprintf(buf,"\tLN:%d",fastaFile.vnSequenceLengths[i]);
s += buf;
if ( !sAS.empty() ) {
sprintf(buf,"\tAS:%s",sAS.c_str());
s += buf;
}
if ( !sUR.empty() ) {
sprintf(buf,"\tUR:%s",sUR.c_str());
s += buf;
}
sprintf(buf,"\tM5:%s",fastaFile.vsMD5sums[i].c_str());
s += buf;
if ( !sSP.empty() ) {
sprintf(buf,"\tSP:%s",sSP.c_str());
s += buf;
}
outHeader.addHeaderLine(s.c_str());
}*/
for(unsigned int i=0; i < vsRGHeaders.size(); ++i) {
samHeader.addHeaderLine(vsRGHeaders[i].c_str());
}
for(unsigned int i=0; i < vsPGHeaders.size(); ++i) {
samHeader.addHeaderLine(vsPGHeaders[i].c_str());
}
samOut.WriteHeader(samHeader);
gpLogger->write_log("Adding %d HD, %d RG, and %d PG headers",vsHDHeaders.size(), vsRGHeaders.size(), vsPGHeaders.size());
gpLogger->write_log("Finished writing output headers");
// parse RG tag and get RG ID to append
std::string sRGID;
if ( ! vsRGHeaders.empty() ) {
std::vector<std::string> tokens;
FastaFile::tokenizeString( vsRGHeaders[0].c_str(), tokens );
for(unsigned int i=0; i < tokens.size(); ++i) {
if ( tokens[i].find("ID:") == 0 ) {
sRGID = tokens[i].substr(3);
}
}
}
gpLogger->write_log("Writing output BAM file");
SamRecord samRecord;
while (samIn.ReadRecord(samHeader, samRecord) == true) {
if ( !sRGID.empty() ) {
if ( samRecord.addTag("RG",'Z',sRGID.c_str()) == false ) {
gpLogger->error("Failed to add a RG tag %s",sRGID.c_str());
}
// temporary code added
if ( strncmp(samRecord.getReadName(),"seqcore_",8) == 0 ) {
char buf[1024];
sprintf(buf,"UM%s",samRecord.getReadName()+8);
samRecord.setReadName(buf);
}
}
samOut.WriteRecord(samHeader, samRecord);
//if ( samIn.GetCurrentRecordCount() == 1000 ) break;
}
samOut.Close();
gpLogger->write_log("Successfully written %d records",samIn.GetCurrentRecordCount());
delete gpLogger;
return 0;
}
bool Recab::processReadApplyTable(SamRecord& samRecord)
{
static BaseData data;
static std::string readGroup;
static std::string aligTypes;
int seqLen = samRecord.getReadLength();
uint16_t flag = samRecord.getFlag();
// Check if the flag contains an exclude.
if((flag & myIntApplyExcludeFlags) != 0)
{
// Do not apply the recalibration table to this read.
++myNumApplySkipped;
return(false);
}
++myNumApplyReads;
readGroup = samRecord.getString("RG").c_str();
// Look for the read group in the map.
// TODO - extra string constructor??
RgInsertReturn insertRet =
myRg2Id.insert(std::pair<std::string, uint16_t>(readGroup, 0));
if(insertRet.second == true)
{
// New element inserted.
insertRet.first->second = myId2Rg.size();
myId2Rg.push_back(readGroup);
}
data.rgid = insertRet.first->second;
if(!myQField.IsEmpty())
{
// Check if there is an old quality.
const String* oldQPtr =
samRecord.getStringTag(myQField.c_str());
if((oldQPtr != NULL) && (oldQPtr->Length() == seqLen))
{
// There is an old quality, so use that.
myQualityStrings.oldq = oldQPtr->c_str();
}
else
{
myQualityStrings.oldq = samRecord.getQuality();
}
}
else
{
myQualityStrings.oldq = samRecord.getQuality();
}
if(myQualityStrings.oldq.length() != (unsigned int)seqLen)
{
Logger::gLogger->warning("Quality is not the correct length, so skipping recalibration on that record.");
return(false);
}
myQualityStrings.newq.resize(seqLen);
////////////////
////// iterate sequence
////////////////
int32_t seqPos = 0;
int seqIncr = 1;
bool reverse;
if(SamFlag::isReverse(flag))
{
reverse = true;
seqPos = seqLen - 1;
seqIncr = -1;
}
else
reverse = false;
// Check which read - this will be the same for all positions, so
// do this outside of the smaller loop.
if(!SamFlag::isPaired(flag) || SamFlag::isFirstFragment(flag))
// Mark as first if it is not paired or if it is the
// first in the pair.
data.read = 0;
else
data.read = 1;
// Set unsetbase for curBase.
// This will be used for the prebase of cycle 0.
data.curBase = 'K';
for (data.cycle = 0; data.cycle < seqLen; data.cycle++, seqPos += seqIncr)
{
// Set the preBase to the previous cycle's current base.
// For cycle 0, curBase was set to a default value.
data.preBase = data.curBase;
// Get the current base.
data.curBase = samRecord.getSequence(seqPos);
if(reverse)
{
// Complement the current base.
data.curBase =
BaseAsciiMap::base2complement[(unsigned int)(data.curBase)];
}
// Get quality
data.qual =
BaseUtilities::getPhredBaseQuality(myQualityStrings.oldq[seqPos]);
// skip bases with quality below the minimum set.
if(data.qual < myMinBaseQual)
{
myQualityStrings.newq[seqPos] = myQualityStrings.oldq[seqPos];
continue;
}
// Update quality score
uint8_t qemp = hasherrormodel.getQemp(data);
qemp = mySqueeze.getQualCharFromQemp(qemp);
if(qemp > myMaxBaseQualChar)
{
qemp = myMaxBaseQualChar;
}
myQualityStrings.newq[seqPos] = qemp;
}
if(!myStoreQualTag.IsEmpty())
{
samRecord.addTag(myStoreQualTag, 'Z', myQualityStrings.oldq.c_str());
}
samRecord.setQuality(myQualityStrings.newq.c_str());
return true;
}
// add a readgroup tag to each SamRecord
void addReadGroupTag(SamRecord& record, ReadGroup& rg) {
if ( record.addTag("RG",'Z',rg.s_id.c_str()) == false ) {
Logger::gLogger->error("Failed to add readgroup tag %s",rg.s_id.c_str());
}
}
