void modify::modifyTags()
{
assert(samIn.OpenForRead(myFilename.c_str()));
// Read the sam header.
assert(samIn.ReadHeader(samHeader));
SamFile samOut;
SamFile bamOut;
std::string inputType = myFilename.substr(myFilename.find_last_of('.'));
std::string outFileBase = "results/updateTagFrom";
if(inputType == ".bam")
{
outFileBase += "Bam";
}
else
{
outFileBase += "Sam";
}
std::string outFile = outFileBase + ".sam";
assert(samOut.OpenForWrite(outFile.c_str()));
outFile = outFileBase + ".bam";
assert(bamOut.OpenForWrite(outFile.c_str()));
assert(samOut.WriteHeader(samHeader));
assert(bamOut.WriteHeader(samHeader));
int count = 0;
// Read the records.
while(samIn.ReadRecord(samHeader, samRecord))
{
if(count == 0)
{
assert(samRecord.rmTag("MD", 'Z'));
}
else if(count == 2)
{
assert(samRecord.rmTags("XT:A;MD:Z;AB:c;NM:i"));
}
else if(count == 4)
{
assert(samRecord.rmTags("MD:Z,AB:c,NM:i"));
}
assert(bamOut.WriteRecord(samHeader, samRecord));
assert(samOut.WriteRecord(samHeader, samRecord));
++count;
}
}
int Dedup_LowMem::execute(int argc, char** argv)
{
/* --------------------------------
* process the arguments
* -------------------------------*/
String inFile, outFile, logFile;
myDoRecab = false;
bool removeFlag = false;
bool verboseFlag = false;
myForceFlag = false;
myNumMissingMate = 0;
myMinQual = DEFAULT_MIN_QUAL;
String excludeFlags = "0xB04";
uint16_t intExcludeFlags = 0;
bool noeof = false;
bool params = false;
LongParamContainer parameters;
parameters.addGroup("Required Parameters");
parameters.addString("in", &inFile);
parameters.addString("out", &outFile);
parameters.addGroup("Optional Parameters");
parameters.addInt("minQual", & myMinQual);
parameters.addString("log", &logFile);
parameters.addBool("oneChrom", &myOneChrom);
parameters.addBool("recab", &myDoRecab);
parameters.addBool("rmDups", &removeFlag);
parameters.addBool("force", &myForceFlag);
parameters.addString("excludeFlags", &excludeFlags);
parameters.addBool("verbose", &verboseFlag);
parameters.addBool("noeof", &noeof);
parameters.addBool("params", ¶ms);
parameters.addPhoneHome(VERSION);
myRecab.addRecabSpecificParameters(parameters);
ParameterList inputParameters;
inputParameters.Add(new LongParameters ("Input Parameters",
parameters.getLongParameterList()));
// parameters start at index 2 rather than 1.
inputParameters.Read(argc, argv, 2);
// 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);
}
if(inFile.IsEmpty())
{
printUsage(std::cerr);
inputParameters.Status();
std::cerr << "Specify an input file" << std::endl;
return EXIT_FAILURE;
}
if(outFile.IsEmpty())
{
printUsage(std::cerr);
inputParameters.Status();
std::cerr << "Specify an output file" << std::endl;
return EXIT_FAILURE;
}
intExcludeFlags = excludeFlags.AsInteger();
if(myForceFlag && SamFlag::isDuplicate(intExcludeFlags))
{
printUsage(std::cerr);
inputParameters.Status();
std::cerr << "Cannot specify --force and Duplicate in the excludeFlags. Since --force indicates to override"
<< " previous duplicate setting and the excludeFlags says to skip those, you can't do both.\n";
return EXIT_FAILURE;
}
if(!SamFlag::isSecondary(intExcludeFlags))
{
printUsage(std::cerr);
inputParameters.Status();
std::cerr << "ERROR: Secondary reads must be excluded, edit --excludeFlags to include 0x0100\n";
return EXIT_FAILURE;
}
if(!(intExcludeFlags & SamFlag::SUPPLEMENTARY_ALIGNMENT))
{
printUsage(std::cerr);
inputParameters.Status();
std::cerr << "ERROR: Supplementary reads must be excluded, edit --excludeFlags to include 0x0800\n";
return EXIT_FAILURE;
}
if(logFile.IsEmpty())
{
logFile = outFile + ".log";
}
if(myDoRecab)
{
int status = myRecab.processRecabParam();
if(status != 0)
{
inputParameters.Status();
return(status);
}
}
if(params)
{
inputParameters.Status();
}
Logger::gLogger = new Logger(logFile.c_str(), verboseFlag);
/* -------------------------------------------------------------------
* The arguments are processed. Prepare the input BAM file,
* instantiate dedup_LowMem, and construct the read group library map
* ------------------------------------------------------------------*/
SamFile samIn;
samIn.OpenForRead(inFile.c_str());
// If the file isn't sorted it will throw an exception.
samIn.setSortedValidation(SamFile::COORDINATE);
SamFileHeader header;
samIn.ReadHeader(header);
buildReadGroupLibraryMap(header);
lastReference = -1;
lastCoordinate = -1;
// for keeping some basic statistics
uint32_t recordCount = 0;
uint32_t pairedCount = 0;
uint32_t properPairCount = 0;
uint32_t unmappedCount = 0;
uint32_t reverseCount = 0;
uint32_t qualCheckFailCount = 0;
uint32_t secondaryCount = 0;
uint32_t supplementaryCount = 0;
uint32_t excludedCount = 0;
// Now we start reading records
SamRecord* recordPtr;
SamStatus::Status returnStatus = SamStatus::SUCCESS;
while(returnStatus == SamStatus::SUCCESS)
{
recordPtr = mySamPool.getRecord();
if(recordPtr == NULL)
{
std::cerr << "Failed to allocate enough records\n";
return(-1);
}
if(!samIn.ReadRecord(header, *recordPtr))
{
returnStatus = samIn.GetStatus();
continue;
}
// Take note of properties of this record
int flag = recordPtr->getFlag();
if(SamFlag::isPaired(flag)) ++pairedCount;
if(SamFlag::isProperPair(flag)) ++properPairCount;
if(SamFlag::isReverse(flag)) ++reverseCount;
if(SamFlag::isQCFailure(flag)) ++qualCheckFailCount;
if(SamFlag::isSecondary(flag)) ++secondaryCount;
if(flag & SamFlag::SUPPLEMENTARY_ALIGNMENT) ++supplementaryCount;
if(!SamFlag::isMapped(flag)) ++unmappedCount;
// put the record in the appropriate maps:
// single reads go in myFragmentMap
// paired reads go in myPairedMap
recordCount = samIn.GetCurrentRecordCount();
// if we have moved to a new position, look back at previous reads for duplicates
if (hasPositionChanged(*recordPtr))
{
cleanupPriorReads(recordPtr);
}
// Determine if this read should be checked for duplicates.
if((!SamFlag::isMapped(flag)) || ((flag & intExcludeFlags) != 0))
{
++excludedCount;
// No deduping done on this record, but still build the recab table.
if(myDoRecab)
{
myRecab.processReadBuildTable(*recordPtr);
}
// Nothing more to do with this record, so
// release the pointer.
mySamPool.releaseRecord(recordPtr);
}
else
{
if(SamFlag::isDuplicate(flag) && !myForceFlag)
{
// Error: Marked duplicates, and duplicates aren't excluded.
Logger::gLogger->error("There are records already duplicate marked.");
Logger::gLogger->error("Use -f to clear the duplicate flag and start the dedup_LowMem procedure over");
}
checkDups(*recordPtr, recordCount);
mySamPool.releaseRecord(recordPtr);
}
// let the user know we're not napping
if (verboseFlag && (recordCount % 100000 == 0))
{
Logger::gLogger->writeLog("recordCount=%u singleKeyMap=%u pairedKeyMap=%u, dictSize=%u",
recordCount, myFragmentMap.size(),
myPairedMap.size(),
myMateMap.size());
}
}
// we're finished reading record so clean up the duplicate search and
// close the input file
cleanupPriorReads(NULL);
samIn.Close();
// print some statistics
Logger::gLogger->writeLog("--------------------------------------------------------------------------");
Logger::gLogger->writeLog("SUMMARY STATISTICS OF THE READS");
Logger::gLogger->writeLog("Total number of reads: %u",recordCount);
Logger::gLogger->writeLog("Total number of paired-end reads: %u",
pairedCount);
Logger::gLogger->writeLog("Total number of properly paired reads: %u",
properPairCount);
Logger::gLogger->writeLog("Total number of unmapped reads: %u",
unmappedCount);
Logger::gLogger->writeLog("Total number of reverse strand mapped reads: %u",
reverseCount);
Logger::gLogger->writeLog("Total number of QC-failed reads: %u",
qualCheckFailCount);
Logger::gLogger->writeLog("Total number of secondary reads: %u",
secondaryCount);
Logger::gLogger->writeLog("Total number of supplementary reads: %u",
supplementaryCount);
Logger::gLogger->writeLog("Size of singleKeyMap (must be zero): %u",
myFragmentMap.size());
Logger::gLogger->writeLog("Size of pairedKeyMap (must be zero): %u",
myPairedMap.size());
Logger::gLogger->writeLog("Total number of missing mates: %u",
myNumMissingMate);
Logger::gLogger->writeLog("Total number of reads excluded from duplicate checking: %u",
excludedCount);
Logger::gLogger->writeLog("--------------------------------------------------------------------------");
Logger::gLogger->writeLog("Sorting the indices of %d duplicated records",
myDupList.size());
// sort the indices of duplicate records
std::sort(myDupList.begin(), myDupList.end(),
std::less<uint32_t> ());
// get ready to write the output file by making a second pass
// through the input file
samIn.OpenForRead(inFile.c_str());
samIn.ReadHeader(header);
SamFile samOut;
samOut.OpenForWrite(outFile.c_str());
samOut.WriteHeader(header);
// If we are recalibrating, output the model information.
if(myDoRecab)
{
myRecab.modelFitPrediction(outFile);
}
// an iterator to run through the duplicate indices
int currentDupIndex = 0;
bool moreDups = !myDupList.empty();
// let the user know what we're doing
Logger::gLogger->writeLog("\nWriting %s", outFile.c_str());
// count the duplicate records as a check
uint32_t singleDuplicates(0), pairedDuplicates(0);
// start reading records and writing them out
SamRecord record;
while(samIn.ReadRecord(header, record))
{
uint32_t currentIndex = samIn.GetCurrentRecordCount();
bool foundDup = moreDups &&
(currentIndex == myDupList[currentDupIndex]);
// modify the duplicate flag and write out the record,
// if it's appropriate
int flag = record.getFlag();
if (foundDup)
{
// this record is a duplicate, so mark it.
record.setFlag( flag | 0x400 );
currentDupIndex++;
// increment duplicate counters to verify we found them all
if ( ( ( flag & 0x0001 ) == 0 ) || ( flag & 0x0008 ) )
{ // unpaired or mate unmapped
singleDuplicates++;
}
else
{
pairedDuplicates++;
}
// recalibrate if necessary.
if(myDoRecab)
{
myRecab.processReadApplyTable(record);
}
// write the record if we are not removing duplicates
if (!removeFlag ) samOut.WriteRecord(header, record);
}
else
{
if(myForceFlag)
{
// this is not a duplicate we've identified but we want to
// remove any duplicate marking
record.setFlag( flag & 0xfffffbff ); // unmark duplicate
}
// Not a duplicate, so recalibrate if necessary.
if(myDoRecab)
{
myRecab.processReadApplyTable(record);
}
samOut.WriteRecord(header, record);
}
// Let the user know we're still here
if (verboseFlag && (currentIndex % 100000 == 0)) {
Logger::gLogger->writeLog("recordCount=%u", currentIndex);
}
}
// We're done. Close the files and print triumphant messages.
samIn.Close();
samOut.Close();
Logger::gLogger->writeLog("Successfully %s %u unpaired and %u paired duplicate reads",
removeFlag ? "removed" : "marked" ,
singleDuplicates,
pairedDuplicates/2);
Logger::gLogger->writeLog("\nDedup_LowMem complete!");
return 0;
}
int Revert::execute(int argc, char **argv)
{
// Extract command line arguments.
String inFile = "";
String outFile = "";
bool cigar = false;
bool qual = false;
bool noeof = false;
bool params = false;
bool rmBQ = false;
String rmTags = "";
myKeepTags = false;
ParameterList inputParameters;
BEGIN_LONG_PARAMETERS(longParameterList)
LONG_STRINGPARAMETER("in", &inFile)
LONG_STRINGPARAMETER("out", &outFile)
LONG_PARAMETER("cigar", &cigar)
LONG_PARAMETER("qual", &qual)
LONG_PARAMETER("keepTags", &myKeepTags)
LONG_PARAMETER("rmBQ", &rmBQ)
LONG_STRINGPARAMETER("rmTags", &rmTags)
LONG_PARAMETER("noeof", &noeof)
LONG_PARAMETER("params", ¶ms)
LONG_PHONEHOME(VERSION)
END_LONG_PARAMETERS();
inputParameters.Add(new LongParameters ("Input Parameters",
longParameterList));
// parameters start at index 2 rather than 1.
inputParameters.Read(argc, argv, 2);
// 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);
}
// Check to see if the in file was specified, if not, report an error.
if(inFile == "")
{
usage();
inputParameters.Status();
// In file was not specified but it is mandatory.
std::cerr << "--in is a mandatory argument, "
<< "but was not specified" << std::endl;
return(-1);
}
if(outFile == "")
{
usage();
inputParameters.Status();
// In file was not specified but it is mandatory.
std::cerr << "--out is a mandatory argument, "
<< "but was not specified" << std::endl;
return(-1);
}
if(params)
{
inputParameters.Status();
}
// Open the input file for reading.
SamFile samIn;
samIn.OpenForRead(inFile);
// Open the output file for writing.
SamFile samOut;
samOut.OpenForWrite(outFile);
// Read the sam header.
SamFileHeader samHeader;
samIn.ReadHeader(samHeader);
// Write the sam header.
samOut.WriteHeader(samHeader);
SamRecord samRecord;
// Set returnStatus to success. It will be changed to the
// failure reason if any of the writes or updates fail.
SamStatus::Status returnStatus = SamStatus::SUCCESS;
// Keep reading records until ReadRecord returns false.
while(samIn.ReadRecord(samHeader, samRecord))
{
// Update the cigar & position.
if(cigar)
{
if(!updateCigar(samRecord))
{
// Failed to update the cigar & position.
fprintf(stderr, "%s\n", samIn.GetStatusMessage());
returnStatus = samIn.GetStatus();
}
}
if(qual)
{
if(!updateQual(samRecord))
{
// Failed to update the quality.
fprintf(stderr, "%s\n", samIn.GetStatusMessage());
returnStatus = samIn.GetStatus();
}
}
if(rmBQ)
{
if(!removeBQ(samRecord))
{
// Failed to remove BQ.
fprintf(stderr, "%s\n", samIn.GetStatusMessage());
returnStatus = samIn.GetStatus();
}
}
if(rmTags != "")
{
if(!samRecord.rmTags(rmTags.c_str()))
{
// Failed to remove the specified tags.
fprintf(stderr, "%s\n", samIn.GetStatusMessage());
returnStatus = samIn.GetStatus();
}
}
// Successfully read a record from the file, so write it.
if(!samOut.WriteRecord(samHeader, samRecord))
{
// Failed to write a record.
fprintf(stderr, "%s\n", samOut.GetStatusMessage());
returnStatus = samOut.GetStatus();
}
}
std::cerr << std::endl << "Number of records read = " <<
samIn.GetCurrentRecordCount() << std::endl;
std::cerr << "Number of records written = " <<
samOut.GetCurrentRecordCount() << std::endl;
// Since the reads were successful, return the status based
// on the status of the writes. If any failed, return
// their failure status.
return(returnStatus);
}
int Convert::execute(int argc, char **argv)
{
// Extract command line arguments.
String inFile = "";
String outFile = "";
String refFile = "";
bool lshift = false;
bool noeof = false;
bool params = false;
bool useBases = false;
bool useEquals = false;
bool useOrigSeq = false;
bool recover = false;
ParameterList inputParameters;
BEGIN_LONG_PARAMETERS(longParameterList)
LONG_STRINGPARAMETER("in", &inFile)
LONG_STRINGPARAMETER("out", &outFile)
LONG_STRINGPARAMETER("refFile", &refFile)
LONG_PARAMETER("lshift", &lshift)
LONG_PARAMETER("noeof", &noeof)
LONG_PARAMETER("recover", &recover)
LONG_PARAMETER("params", ¶ms)
LONG_PARAMETER_GROUP("SequenceConversion")
EXCLUSIVE_PARAMETER("useBases", &useBases)
EXCLUSIVE_PARAMETER("useEquals", &useEquals)
EXCLUSIVE_PARAMETER("useOrigSeq", &useOrigSeq)
LONG_PHONEHOME(VERSION)
END_LONG_PARAMETERS();
inputParameters.Add(new LongParameters ("Input Parameters",
longParameterList));
// parameters start at index 2 rather than 1.
inputParameters.Read(argc, argv, 2);
// 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);
}
// Check to see if the in file was specified, if not, report an error.
if(inFile == "")
{
printUsage(std::cerr);
inputParameters.Status();
// In file was not specified but it is mandatory.
std::cerr << "--in is a mandatory argument, "
<< "but was not specified" << std::endl;
return(-1);
}
if(outFile == "")
{
printUsage(std::cerr);
inputParameters.Status();
// In file was not specified but it is mandatory.
std::cerr << "--out is a mandatory argument, "
<< "but was not specified" << std::endl;
return(-1);
}
// Check to see if the ref file was specified.
// Open the reference.
GenomeSequence* refPtr = NULL;
if(refFile != "")
{
refPtr = new GenomeSequence(refFile);
}
SamRecord::SequenceTranslation translation;
if((useBases) && (refPtr != NULL))
{
translation = SamRecord::BASES;
}
else if((useEquals) && (refPtr != NULL))
{
translation = SamRecord::EQUAL;
}
else
{
useOrigSeq = true;
translation = SamRecord::NONE;
}
if(params)
{
inputParameters.Status();
}
// Open the input file for reading.
SamFile samIn;
if(recover) samIn.setAttemptRecovery(true);
samIn.OpenForRead(inFile);
// Open the output file for writing.
SamFile samOut;
samOut.OpenForWrite(outFile);
samOut.SetWriteSequenceTranslation(translation);
samOut.SetReference(refPtr);
// Read the sam header.
SamFileHeader samHeader;
samIn.ReadHeader(samHeader);
// Write the sam header.
samOut.WriteHeader(samHeader);
SamRecord samRecord;
// Set returnStatus to success. It will be changed
// to the failure reason if any of the writes fail.
SamStatus::Status returnStatus = SamStatus::SUCCESS;
while(1) {
try {
// Keep reading records until ReadRecord returns false.
while(samIn.ReadRecord(samHeader, samRecord))
{
// left shift if necessary.
if(lshift)
{
samRecord.shiftIndelsLeft();
}
// Successfully read a record from the file, so write it.
if(!samOut.WriteRecord(samHeader, samRecord))
{
// Failed to write a record.
fprintf(stderr, "%s\n", samOut.GetStatusMessage());
returnStatus = samOut.GetStatus();
}
}
break;
} catch (std::runtime_error e) {
std::cerr << "Caught runtime error: " << e.what() << "\n";
if(!recover) {
std::cerr << "Corrupted BAM file detected - consider using --recover option.\n";
break;
}
std::cerr << "Attempting to resync at next good BGZF block and BAM record.\n";
// XXX need to resync SamFile stream here
bool rc = samIn.attemptRecoverySync(checkSignature, SIGNATURE_LENGTH);
if(rc) {
std::cerr << "Successful resync - some data lost.\n";
continue; // succeeded
}
std::cerr << "Failed to re-sync on data stream.\n";
break; // failed to resync
}
}
std::cerr << std::endl << "Number of records read = " <<
samIn.GetCurrentRecordCount() << std::endl;
std::cerr << "Number of records written = " <<
samOut.GetCurrentRecordCount() << std::endl;
if(refPtr != NULL)
{
delete(refPtr);
}
// Since the reads were successful, return the status based
// on the status of the writes. If any failed, return
// their failure status.
return(returnStatus);
}
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 BamProcessor::init (const ContalignParams& p)
{
read_cnt_ = proc_cnt_ = toolongs_ = unaligned_cnt_ = fail_cnt_ = nomd_cnt_ = realigned_cnt_ = modified_cnt_ = pos_adjusted_cnt_ = 0;
log_diff_ = log_matr_ = log_base_ = false;
p_ = &p;
if (!*p.inbam ())
ers << "Input file name not specified" << Throw;
limit_ = p.limit ();
skip_ = p.skip ();
infile_.OpenForRead (p.inbam ());
if (!infile_.IsOpen ())
ers << p.inbam () << ThrowEx (FileNotFoundRerror);
bool index_ok = false;
if (*p.bamidx ())
{
index_ok = infile_.ReadBamIndex (p.bamidx ());
if (!index_ok)
warn << "Unable to open specified BAM index: " << p.bamidx () << ". Default index will be attempted" << std::endl;
}
if (!index_ok)
{
try
{
index_ok = infile_.ReadBamIndex ();
}
catch (std::exception& e)
{
// for some reason not converted into return status by libStatGen
}
if (!index_ok)
warn << "Unable to open default BAM index for " << p.inbam () << std::endl;
}
if (*p.refname () || p.refno () != -1)
{
if (!index_ok)
ers << "Reference section specified, but the BAM index could not be open." << Throw;
if (*p.refname ())
{
if (p.endpos () != 0)
{
infile_.SetReadSection (p.refname (), p.begpos (), p.endpos ());
info << "Read section set : " << p.refname () << ": " << p.begpos () << "-" << p.endpos () << std::endl;
}
else
{
infile_.SetReadSection (p.refname ());
info << "Read section set : " << p.refname () << std::endl;
}
}
else
{
if (p.endpos () != 0)
{
info << "Read section set : ref# " << p.refno () << ": " << p.begpos () << "-" << p.endpos () << std::endl;
infile_.SetReadSection (p.refno (), p.begpos (), p.endpos ());
}
else
{
info << "Read section set : ref# " << p.refno () << std::endl;
infile_.SetReadSection (p.refno ());
}
}
}
if (*p.outbam ())
{
if (!p.overwrite () && file_exists (p.outbam ()))
ers << "Output file " << p.outbam () << " exists. Use --ov key to allow overwriting" << Throw;
outfile_.OpenForWrite (p.outbam ());
if (!outfile_.IsOpen ())
ers << "Unable to open output file " << p.outbam () << std::endl;
}
if (*p.logfname ())
{
if (!p.overwrite () && file_exists (p.logfname ()))
ers << "Log file " << p.logfname () << " exists. Use --ov key to allow overwriting" << Throw;
logfile_.open (p.logfname (), std::fstream::out);
if (!logfile_.is_open ())
ers << "Unable to open log file " << p.logfname () << std::endl;
time_t t = time (NULL);
logfile_ << "Context-aware realigner log\nStarted at " << asctime (localtime (&t)) << "\nParameters:\n";
logfile_ << *(p.parameters_);
logfile_ << std::endl;
log_base_ = p.logging ("base");
log_diff_ = p.logging ("diff");
log_matr_ = p.logging ("matr");
}
band_width_ = p.bwid ();
switch (p.algo ())
{
case ContalignParams::TEMPL:
{
matrix_.configure (genstr::nucleotides.symbols (), genstr::nucleotides.size (), genstr::NegUnitaryMatrix <int, 4>().values ());
gap_cost_.configure (p.gip (), p.gep ());
taligner_.configure (&matrix_, &gap_cost_, &gap_cost_, &genstr::nn2num, &genstr::nn2num);
}
break;
case ContalignParams::PLAIN:
{
batches_.reset (max_batch_no_);
aligner_.init (MAX_SEQ_LEN, MAX_SEQ_LEN*MAX_BAND_WIDTH, p.gip (), p.gep (), p.mat (), -p.mis ());
if (log_matr_)
aligner_.set_log (logfile_);
if (p.debug () > 5)
aligner_.set_trace (true);
}
break;
case ContalignParams::POLY:
{
batches_.reset (max_batch_no_);
contalign_.init (MAX_SEQ_LEN, MAX_RSEQ_LEN, MAX_SEQ_LEN*MAX_BAND_WIDTH, p.gip (), p.gep (), p.mat (), -p.mis ());
if (log_matr_)
contalign_.set_log (logfile_);
if (p.debug () > 5)
contalign_.set_trace (true);
}
break;
default:
{
ers << "Alignment algorithm " << p.algostr () << " not yet supported" << Throw;
}
}
timer_.reset (DEFAULT_REPORT_IVAL, 1);
return true;
}
// main function
int TrimBam::execute(int argc, char ** argv)
{
SamFile samIn;
SamFile samOut;
int numTrimBaseL = 0;
int numTrimBaseR = 0;
bool noeof = false;
bool ignoreStrand = false;
bool noPhoneHome = false;
std::string inName = "";
std::string outName = "";
if ( argc < 5 ) {
usage();
std::cerr << "ERROR: Incorrect number of parameters specified\n";
return(-1);
}
inName = argv[2];
outName = argv[3];
static struct option getopt_long_options[] = {
// Input options
{ "left", required_argument, NULL, 'L'},
{ "right", required_argument, NULL, 'R'},
{ "ignoreStrand", no_argument, NULL, 'i'},
{ "noeof", no_argument, NULL, 'n'},
{ "noPhoneHome", no_argument, NULL, 'p'},
{ "nophonehome", no_argument, NULL, 'P'},
{ "phoneHomeThinning", required_argument, NULL, 't'},
{ "phonehomethinning", required_argument, NULL, 'T'},
{ NULL, 0, NULL, 0 },
};
int argIndex = 4;
if(argv[argIndex][0] != '-')
{
// This is the number of bases to trim off both sides
// so convert to a number.
numTrimBaseL = atoi(argv[argIndex]);
numTrimBaseR = numTrimBaseL;
++argIndex;
}
int c = 0;
int n_option_index = 0;
// Process any additional parameters
while ( ( c = getopt_long(argc, argv,
"L:R:in", getopt_long_options, &n_option_index) )
!= -1 )
{
switch(c)
{
case 'L':
numTrimBaseL = atoi(optarg);
break;
case 'R':
numTrimBaseR = atoi(optarg);
break;
case 'i':
ignoreStrand = true;
break;
case 'n':
noeof = true;
break;
case 'p':
case 'P':
noPhoneHome = true;
break;
case 't':
case 'T':
PhoneHome::allThinning = atoi(optarg);
break;
default:
fprintf(stderr,"ERROR: Unrecognized option %s\n",
getopt_long_options[n_option_index].name);
return(-1);
}
}
if(!noPhoneHome)
{
PhoneHome::checkVersion(getProgramName(), VERSION);
}
if(noeof)
{
// Set that the eof block is not required.
BgzfFileType::setRequireEofBlock(false);
}
if ( ! samIn.OpenForRead(inName.c_str()) ) {
fprintf(stderr, "***Problem opening %s\n",inName.c_str());
return(-1);
}
if(!samOut.OpenForWrite(outName.c_str())) {
fprintf(stderr, "%s\n", samOut.GetStatusMessage());
return(samOut.GetStatus());
}
fprintf(stderr,"Arguments in effect: \n");
fprintf(stderr,"\tInput file : %s\n",inName.c_str());
fprintf(stderr,"\tOutput file : %s\n",outName.c_str());
if(numTrimBaseL == numTrimBaseR)
{
fprintf(stderr,"\t#Bases to trim from each side : %d\n", numTrimBaseL);
}
else
{
fprintf(stderr,"\t#Bases to trim from the left of forward strands : %d\n",
numTrimBaseL);
fprintf(stderr,"\t#Bases to trim from the right of forward strands: %d\n",
numTrimBaseR);
if(!ignoreStrand)
{
// By default, reverse strands are treated the opposite.
fprintf(stderr,"\t#Bases to trim from the left of reverse strands : %d\n",
numTrimBaseR);
fprintf(stderr,"\t#Bases to trim from the right of reverse strands : %d\n",
numTrimBaseL);
}
else
{
// ignore strand, treating forward & reverse strands the same
fprintf(stderr,"\t#Bases to trim from the left of reverse strands : %d\n",
numTrimBaseL);
fprintf(stderr,"\t#Bases to trim from the right of reverse strands : %d\n",
numTrimBaseR);
}
}
// Read the sam header.
SamFileHeader samHeader;
if(!samIn.ReadHeader(samHeader))
{
fprintf(stderr, "%s\n", samIn.GetStatusMessage());
return(samIn.GetStatus());
}
// Write the sam header.
if(!samOut.WriteHeader(samHeader))
{
fprintf(stderr, "%s\n", samOut.GetStatusMessage());
return(samOut.GetStatus());
}
SamRecord samRecord;
char seq[65536];
char qual[65536];
int i, len;
// Keep reading records until ReadRecord returns false.
while(samIn.ReadRecord(samHeader, samRecord)) {
// Successfully read a record from the file, so write it.
strcpy(seq,samRecord.getSequence());
strcpy(qual,samRecord.getQuality());
// Number of bases to trim from the left/right,
// set based on ignoreStrand flag and strand info.
int trimLeft = numTrimBaseL;
int trimRight = numTrimBaseR;
if(!ignoreStrand)
{
if(SamFlag::isReverse(samRecord.getFlag()))
{
// We are reversing the reverse reads,
// so swap the left & right trim counts.
trimRight = numTrimBaseL;
trimLeft = numTrimBaseR;
}
}
len = strlen(seq);
// Do not trim if sequence is '*'
if ( strcmp(seq, "*") != 0 ) {
bool qualValue = true;
if(strcmp(qual, "*") == 0)
{
qualValue = false;
}
int qualLen = strlen(qual);
if ( (qualLen != len) && qualValue ) {
fprintf(stderr,"ERROR: Sequence and Quality have different length\n");
return(-1);
}
if ( len < (trimLeft + trimRight) ) {
// Read Length is less than the total number of bases to trim,
// so trim the entire read.
for(i=0; i < len; ++i) {
seq[i] = 'N';
if ( qualValue ) {
qual[i] = '!';
}
}
}
else
{
// Read Length is larger than the total number of bases to trim,
// so trim from the left, then from the right.
for(i=0; i < trimLeft; ++i)
{
// Trim the bases from the left.
seq[i] = 'N';
if ( qualValue )
{
qual[i] = '!';
}
}
for(i = 0; i < trimRight; i++)
{
seq[len-i-1] = 'N';
if(qualValue)
{
qual[len-i-1] = '!';
}
}
}
samRecord.setSequence(seq);
samRecord.setQuality(qual);
}
if(!samOut.WriteRecord(samHeader, samRecord)) {
// Failed to write a record.
fprintf(stderr, "Failure in writing record %s\n", samOut.GetStatusMessage());
return(-1);
}
}
if(samIn.GetStatus() != SamStatus::NO_MORE_RECS)
{
// Failed to read a record.
fprintf(stderr, "%s\n", samIn.GetStatusMessage());
}
std::cerr << std::endl << "Number of records read = " <<
samIn.GetCurrentRecordCount() << std::endl;
std::cerr << "Number of records written = " <<
samOut.GetCurrentRecordCount() << std::endl;
if(samIn.GetStatus() != SamStatus::NO_MORE_RECS)
{
// Failed reading a record.
return(samIn.GetStatus());
}
// Since the reads were successful, return the status based
samIn.Close();
samOut.Close();
return 0;
}
int WriteRegion::execute(int argc, char **argv)
{
// Extract command line arguments.
String inFile = "";
String outFile = "";
String indexFile = "";
String readName = "";
String bed = "";
myStart = UNSPECIFIED_INT;
myEnd = UNSPECIFIED_INT;
myPrevStart = UNSPECIFIED_INT;
myPrevEnd = UNSPECIFIED_INT;
myRefID = UNSET_REF;
myRefName.Clear();
myPrevRefName.Clear();
myBedRefID = SamReferenceInfo::NO_REF_ID;
bool lshift = false;
bool noeof = false;
bool params = false;
myWithinReg = false;
myWroteReg = false;
ParameterList inputParameters;
BEGIN_LONG_PARAMETERS(longParameterList)
LONG_PARAMETER_GROUP("Required Parameters")
LONG_STRINGPARAMETER("in", &inFile)
LONG_STRINGPARAMETER("out", &outFile)
LONG_PARAMETER_GROUP("Optional Region Parameters")
LONG_STRINGPARAMETER("bamIndex", &indexFile)
LONG_STRINGPARAMETER("refName", &myRefName)
LONG_INTPARAMETER("refID", &myRefID)
LONG_INTPARAMETER("start", &myStart)
LONG_INTPARAMETER("end", &myEnd)
LONG_STRINGPARAMETER("bed", &bed)
LONG_PARAMETER("withinReg", &myWithinReg)
LONG_STRINGPARAMETER("readName", &readName)
LONG_PARAMETER_GROUP("Optional Other Parameters")
LONG_PARAMETER("lshift", &lshift)
LONG_PARAMETER("noeof", &noeof)
LONG_PARAMETER("params", ¶ms)
END_LONG_PARAMETERS();
inputParameters.Add(new LongParameters ("Input Parameters",
longParameterList));
inputParameters.Read(argc-1, &(argv[1]));
// 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);
}
// Check to see if the in file was specified, if not, report an error.
if(inFile == "")
{
usage();
// mandatory argument was not specified.
inputParameters.Status();
std::cerr << "Missing mandatory argument: --in" << std::endl;
return(-1);
}
if(outFile == "")
{
usage();
// mandatory argument was not specified.
inputParameters.Status();
std::cerr << "Missing mandatory argument: --out" << std::endl;
return(-1);
}
if(indexFile == "")
{
// In file was not specified, so set it to the in file
// + ".bai"
indexFile = inFile + ".bai";
}
if(myRefID != UNSET_REF && myRefName.Length() != 0)
{
std::cerr << "Can't specify both refID and refName" << std::endl;
inputParameters.Status();
return(-1);
}
if(myRefID != UNSET_REF && bed.Length() != 0)
{
std::cerr << "Can't specify both refID and bed" << std::endl;
inputParameters.Status();
return(-1);
}
if(myRefName.Length() != 0 && bed.Length() != 0)
{
std::cerr << "Can't specify both refName and bed" << std::endl;
inputParameters.Status();
return(-1);
}
if(!bed.IsEmpty())
{
myBedFile = ifopen(bed, "r");
}
if(params)
{
inputParameters.Status();
}
// Open the file for reading.
mySamIn.OpenForRead(inFile);
// Open the output file for writing.
SamFile samOut;
samOut.OpenForWrite(outFile);
// Open the bam index file for reading if a region was specified.
if((myRefName.Length() != 0) || (myRefID != UNSET_REF) || (myBedFile != NULL))
{
mySamIn.ReadBamIndex(indexFile);
}
// Read & write the sam header.
mySamIn.ReadHeader(mySamHeader);
samOut.WriteHeader(mySamHeader);
// Read the sam records.
SamRecord samRecord;
// Track the status.
int numSectionRecords = 0;
// Set returnStatus to success. It will be changed
// to the failure reason if any of the writes fail.
SamStatus::Status returnStatus = SamStatus::SUCCESS;
while(getNextSection())
{
// Keep reading records until they aren't anymore.
while(mySamIn.ReadRecord(mySamHeader, samRecord))
{
if(!readName.IsEmpty())
{
// Check for readname.
if(strcmp(samRecord.getReadName(), readName.c_str()) != 0)
{
// not a matching read name, so continue to the next record.
continue;
}
}
// Check to see if the read has already been processed.
if(myPrevEnd != UNSPECIFIED_INT)
{
// Because we already know that the bed was sorted,
// we know that the previous section started before
// this one, so if the previous end is greater than
// this record's end position we know that it
// was already written in the previous section.
// Note: can't be equal to the previous end since
// the end range was exclusive, while
// get0BasedAlignmentEnd is inclusive.
// myPrevEnd is reset by getNextSection when a new
// chromosome is hit.
if(samRecord.get0BasedAlignmentEnd() < myPrevEnd)
{
// This record was already written.
continue;
}
}
// Shift left if applicable.
if(lshift)
{
samRecord.shiftIndelsLeft();
}
// Successfully read a record from the file, so write it.
samOut.WriteRecord(mySamHeader, samRecord);
++numSectionRecords;
}
myWroteReg = true;
}
if(myBedFile != NULL)
{
ifclose(myBedFile);
}
std::cerr << "Wrote " << outFile << " with " << numSectionRecords
<< " records.\n";
return(returnStatus);
}
