SamStatus::Status ClipOverlap::readCoordRecord(SamFile& samIn,
SamRecord** recordPtr,
MateMapByCoord& mateMap,
SamCoordOutput* outputBufferPtr)
{
// Null pointer, so get a new pointer.
if(*recordPtr == NULL)
{
*recordPtr = myPool.getRecord();
if(*recordPtr == NULL)
{
// Failed to allocate a new record.
// Try to free up records from the mate map
if(!forceRecordFlush(mateMap, outputBufferPtr))
{
std::cerr << "Failed to flush the output buffer.\n";
return(SamStatus::FAIL_IO);
}
// Try to get a new record, one should have been cleared.
*recordPtr = myPool.getRecord();
if(*recordPtr == NULL)
{
std::cerr << "Failed to allocate any records.\n";
return(SamStatus::FAIL_MEM);
}
}
}
// RecordPtr is set.
if(!samIn.ReadRecord(mySamHeader, **recordPtr))
{
// Nothing to process, so return.
return(samIn.GetStatus());
}
return(SamStatus::SUCCESS);
}
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 Bam2FastQ::execute(int argc, char **argv)
{
// Extract command line arguments.
String inFile = "";
bool readName = false;
String refFile = "";
String firstOut = "";
String secondOut = "";
String unpairedOut = "";
bool interleave = false;
bool noeof = false;
bool gzip = false;
bool params = false;
myOutBase = "";
myNumMateFailures = 0;
myNumPairs = 0;
myNumUnpaired = 0;
mySplitRG = false;
myQField = "";
myNumQualTagErrors = 0;
myReverseComp = true;
myRNPlus = false;
myFirstRNExt = DEFAULT_FIRST_EXT;
mySecondRNExt = DEFAULT_SECOND_EXT;
myCompression = InputFile::DEFAULT;
ParameterList inputParameters;
BEGIN_LONG_PARAMETERS(longParameterList)
LONG_PARAMETER_GROUP("Required Parameters")
LONG_STRINGPARAMETER("in", &inFile)
LONG_PARAMETER_GROUP("Optional Parameters")
LONG_PARAMETER("readName", &readName)
LONG_PARAMETER("splitRG", &mySplitRG)
LONG_STRINGPARAMETER("qualField", &myQField)
LONG_PARAMETER("merge", &interleave)
LONG_STRINGPARAMETER("refFile", &refFile)
LONG_STRINGPARAMETER("firstRNExt", &myFirstRNExt)
LONG_STRINGPARAMETER("secondRNExt", &mySecondRNExt)
LONG_PARAMETER("rnPlus", &myRNPlus)
LONG_PARAMETER("noReverseComp", &myReverseComp)
LONG_PARAMETER("gzip", &gzip)
LONG_PARAMETER("noeof", &noeof)
LONG_PARAMETER("params", ¶ms)
LONG_PARAMETER_GROUP("Optional OutputFile Names")
LONG_STRINGPARAMETER("outBase", &myOutBase)
LONG_STRINGPARAMETER("firstOut", &firstOut)
LONG_STRINGPARAMETER("secondOut", &secondOut)
LONG_STRINGPARAMETER("unpairedOut", &unpairedOut)
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);
}
if(gzip)
{
myCompression = InputFile::GZIP;
}
// 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);
}
// Cannot specify both interleaved & secondOut since secondOut would be N/A.
if(interleave && !secondOut.IsEmpty())
{
usage();
inputParameters.Status();
std::cerr << "ERROR: Cannot specify --merge & --secondOut.\n";
return(-1);
}
// Cannot specify both interleaved & secondOut since secondOut would be N/A.
if(interleave && !secondOut.IsEmpty())
{
usage();
inputParameters.Status();
std::cerr << "ERROR: Cannot specify --merge & --secondOut.\n";
return(-1);
}
// Cannot specify both splitRG & firstOut/secondOut/unpairedOut
// since it needs a different file for each RG.
if(mySplitRG && (!firstOut.IsEmpty() ||
!secondOut.IsEmpty() || !unpairedOut.IsEmpty()))
{
usage();
inputParameters.Status();
std::cerr << "ERROR: Cannot specify --splitRG & --firstOut/--secondOut/--unpairedOut.\n";
std::cerr << "Use --outBase instead.\n";
return(-1);
}
// Cannot specify splitRG & output to stdout.
if(mySplitRG && (myOutBase[0] == '-'))
{
usage();
inputParameters.Status();
std::cerr << "ERROR: Cannot specify --splitRG & write to stdout.\n";
return(-1);
}
// Check to see if the out file was specified, if not, generate it from
// the input filename.
if(myOutBase == "")
{
// Just remove the extension from the input filename.
int extStart = inFile.FastFindLastChar('.');
if(extStart <= 0)
{
myOutBase = inFile;
}
else
{
myOutBase = inFile.Left(extStart);
}
}
if(mySplitRG)
{
std::string fqList = myOutBase.c_str();
fqList += ".list";
myFqList = ifopen(fqList.c_str(), "w");
ifprintf(myFqList, "MERGE_NAME\tFASTQ1\tFASTQ2\tRG\n");
}
// Check to see if the first/second/single-ended were specified and
// if not, set them.
myFirstFileNameExt = "_1.fastq";
mySecondFileNameExt = "_2.fastq";
myUnpairedFileNameExt = ".fastq";
if(interleave)
{
myFirstFileNameExt = "_interleaved.fastq";
myFirstFileNameExt = "_interleaved.fastq";
}
getFileName(firstOut, myFirstFileNameExt);
getFileName(secondOut, mySecondFileNameExt);
getFileName(unpairedOut, myUnpairedFileNameExt);
if(params)
{
inputParameters.Status();
}
// Open the files for reading/writing.
// Open prior to opening the output files,
// so if there is an error, the outputs don't get created.
SamFile samIn;
samIn.OpenForRead(inFile, &mySamHeader);
// Skip non-primary reads.
samIn.SetReadFlags(0, 0x0100);
// Open the output files if not splitting RG
if(!mySplitRG)
{
myUnpairedFile = ifopen(unpairedOut, "w", myCompression);
// Only open the first file if it is different than an already opened file.
if(firstOut != unpairedOut)
{
myFirstFile = ifopen(firstOut, "w", myCompression);
}
else
{
myFirstFile = myUnpairedFile;
}
// If it is interleaved or the 2nd file is not a new name, set it appropriately.
if(interleave || secondOut == firstOut)
{
mySecondFile = myFirstFile;
}
else if(secondOut == unpairedOut)
{
mySecondFile = myUnpairedFile;
}
else
{
mySecondFile = ifopen(secondOut, "w", myCompression);
}
if(myUnpairedFile == NULL)
{
std::cerr << "Failed to open " << unpairedOut
<< " so can't convert bam2FastQ.\n";
return(-1);
}
if(myFirstFile == NULL)
{
std::cerr << "Failed to open " << firstOut
<< " so can't convert bam2FastQ.\n";
return(-1);
}
if(mySecondFile == NULL)
{
std::cerr << "Failed to open " << secondOut
<< " so can't convert bam2FastQ.\n";
return(-1);
}
}
if((readName) || (strcmp(mySamHeader.getSortOrder(), "queryname") == 0))
{
readName = true;
}
else
{
// defaulting to coordinate sorted.
samIn.setSortedValidation(SamFile::COORDINATE);
}
// Setup the '=' translation if the reference was specified.
if(!refFile.IsEmpty())
{
GenomeSequence* refPtr = new GenomeSequence(refFile);
samIn.SetReadSequenceTranslation(SamRecord::BASES);
samIn.SetReference(refPtr);
}
SamRecord* recordPtr;
int16_t samFlag;
SamStatus::Status returnStatus = SamStatus::SUCCESS;
while(returnStatus == SamStatus::SUCCESS)
{
recordPtr = myPool.getRecord();
if(recordPtr == NULL)
{
// Failed to allocate a new record.
throw(std::runtime_error("Failed to allocate a new SAM/BAM record"));
}
if(!samIn.ReadRecord(mySamHeader, *recordPtr))
{
// Failed to read a record.
returnStatus = samIn.GetStatus();
continue;
}
// Have a record. Check to see if it is a pair or unpaired read.
samFlag = recordPtr->getFlag();
if(SamFlag::isPaired(samFlag))
{
if(readName)
{
handlePairedRN(*recordPtr);
}
else
{
handlePairedCoord(*recordPtr);
}
}
else
{
++myNumUnpaired;
writeFastQ(*recordPtr, myUnpairedFile,
myUnpairedFileNameExt);
}
}
// Flush All
cleanUpMateMap(0, true);
if(returnStatus == SamStatus::NO_MORE_RECS)
{
returnStatus = SamStatus::SUCCESS;
}
samIn.Close();
closeFiles();
// Output the results
std::cerr << "\nFound " << myNumPairs << " read pairs.\n";
std::cerr << "Found " << myNumUnpaired << " unpaired reads.\n";
if(myNumMateFailures != 0)
{
std::cerr << "Failed to find mates for " << myNumMateFailures
<< " reads, so they were written as unpaired\n"
<< " (not included in either of the above counts).\n";
}
if(myNumQualTagErrors != 0)
{
std::cerr << myNumQualTagErrors << " records did not have tag "
<< myQField.c_str() << " or it was invalid, so the quality field was used for those records.\n";
}
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;
}
SamStatus::Status ClipOverlap::handleSortedByReadName(SamFile& samIn,
SamFile* samOutPtr)
{
// Set returnStatus to success. It will be changed
// to the failure reason if any of the writes fail.
SamStatus::Status returnStatus = SamStatus::SUCCESS;
// Read the sam records.
SamRecord* prevSamRecord = NULL;
SamRecord* samRecord = new SamRecord;
SamRecord* tmpRecord = new SamRecord;
if((samRecord == NULL) || (tmpRecord == NULL))
{
std::cerr << "Failed to allocate a SamRecord, so exit.\n";
return(SamStatus::FAIL_MEM);
}
// Keep reading records until ReadRecord returns false.
while(samIn.ReadRecord(mySamHeader, *samRecord))
{
int16_t flag = samRecord->getFlag();
if((flag & myIntExcludeFlags) != 0)
{
// This read should not be checked for overlaps.
// Check if there is a previous SamRecord.
if(prevSamRecord != NULL)
{
// There is a previous record.
// If it has a different read name, write it.
if(strcmp(samRecord->getReadName(),
prevSamRecord->getReadName()) != 0)
{
// Different read name, so write the previous record.
if((samOutPtr != NULL) && !myOverlapsOnly)
{
if(!samOutPtr->WriteRecord(mySamHeader, *prevSamRecord))
{
// Failed to write a record.
fprintf(stderr, "%s\n", samOutPtr->GetStatusMessage());
returnStatus = samOutPtr->GetStatus();
}
}
// Clear the previous record info.
tmpRecord = prevSamRecord;
prevSamRecord = NULL;
}
// If it has the same read name, leave it in case there is another read with that name
}
// This record is not being checked for overlaps, so just write it and continue
if((samOutPtr != NULL) && !myOverlapsOnly)
{
if(!samOutPtr->WriteRecord(mySamHeader, *samRecord))
{
// Failed to write a record.
fprintf(stderr, "%s\n", samOutPtr->GetStatusMessage());
returnStatus = samOutPtr->GetStatus();
}
}
continue;
}
if(prevSamRecord == NULL)
{
// Nothing to compare this record to, so set this
// record to the previous, and the next record.
prevSamRecord = samRecord;
samRecord = tmpRecord;
tmpRecord = NULL;
continue;
}
// Check if the read name matches the previous read name.
if(strcmp(samRecord->getReadName(),
prevSamRecord->getReadName()) == 0)
{
bool overlap = false;
// Same Read Name, so check clipping.
OverlapHandler::OverlapInfo prevClipInfo =
myOverlapHandler->getOverlapInfo(*prevSamRecord);
OverlapHandler::OverlapInfo curClipInfo =
myOverlapHandler->getOverlapInfo(*samRecord);
// If either indicate a complete clipping, clip both.
if((prevClipInfo == OverlapHandler::NO_OVERLAP_WRONG_ORIENT) ||
(curClipInfo == OverlapHandler::NO_OVERLAP_WRONG_ORIENT))
{
overlap = true;
myOverlapHandler->handleNoOverlapWrongOrientation(*prevSamRecord);
// Don't update stats since this is the 2nd in the pair
myOverlapHandler->handleNoOverlapWrongOrientation(*samRecord,
false);
}
else if((prevClipInfo == OverlapHandler::OVERLAP) ||
(prevClipInfo == OverlapHandler::SAME_START))
{
// The previous read starts at or before the current one.
overlap = true;
myOverlapHandler->handleOverlapPair(*prevSamRecord,
*samRecord);
}
else if(curClipInfo == OverlapHandler::OVERLAP)
{
// The current read starts before the previous one.
overlap = true;
myOverlapHandler->handleOverlapPair(*samRecord,
*prevSamRecord);
}
// Found a read pair, so write both records if:
// 1) output file is specified
// AND
// 2a) all records should be written
// OR
// 2b) the pair overlaps
if((samOutPtr != NULL) && (!myOverlapsOnly || overlap))
{
if(!samOutPtr->WriteRecord(mySamHeader, *prevSamRecord))
{
// Failed to write a record.
fprintf(stderr, "%s\n", samOutPtr->GetStatusMessage());
returnStatus = samOutPtr->GetStatus();
}
if(!samOutPtr->WriteRecord(mySamHeader, *samRecord))
{
// Failed to write a record.
fprintf(stderr, "%s\n", samOutPtr->GetStatusMessage());
returnStatus = samOutPtr->GetStatus();
}
}
// Setup for the next read with no previous.
tmpRecord = prevSamRecord;
prevSamRecord = NULL;
}
else
{
// Read name does not match, so write the previous record
// if we are writing all records.
if((samOutPtr != NULL) && !myOverlapsOnly)
{
if(!samOutPtr->WriteRecord(mySamHeader, *prevSamRecord))
{
// Failed to write a record.
fprintf(stderr, "%s\n", samOutPtr->GetStatusMessage());
returnStatus = samOutPtr->GetStatus();
}
}
// Store this record as the previous.
tmpRecord = prevSamRecord;
prevSamRecord = samRecord;
samRecord = tmpRecord;
tmpRecord = NULL;
}
}
// Write the previous record if there is one.
if((samOutPtr != NULL) && (prevSamRecord != NULL) && !myOverlapsOnly)
{
if(!samOutPtr->WriteRecord(mySamHeader, *prevSamRecord))
{
// Failed to write a record.
fprintf(stderr, "%s\n", samOutPtr->GetStatusMessage());
returnStatus = samOutPtr->GetStatus();
}
delete prevSamRecord;
}
if(samRecord != NULL)
{
delete samRecord;
}
if(tmpRecord != NULL)
{
delete tmpRecord;
}
if(samIn.GetStatus() != SamStatus::NO_MORE_RECS)
{
return(samIn.GetStatus());
}
return(returnStatus);
}
int Stats::execute(int argc, char **argv)
{
// Extract command line arguments.
String inFile = "";
String indexFile = "";
bool basic = false;
bool noeof = false;
bool params = false;
bool qual = false;
bool phred = false;
int maxNumReads = -1;
bool unmapped = false;
String pBaseQC = "";
String cBaseQC = "";
String regionList = "";
int excludeFlags = 0;
int requiredFlags = 0;
bool withinRegion = false;
int minMapQual = 0;
String dbsnp = "";
PosList *dbsnpListPtr = NULL;
bool baseSum = false;
int bufferSize = PileupHelper::DEFAULT_WINDOW_SIZE;
ParameterList inputParameters;
BEGIN_LONG_PARAMETERS(longParameterList)
LONG_PARAMETER_GROUP("Required Parameters")
LONG_STRINGPARAMETER("in", &inFile)
LONG_PARAMETER_GROUP("Types of Statistics")
LONG_PARAMETER("basic", &basic)
LONG_PARAMETER("qual", &qual)
LONG_PARAMETER("phred", &phred)
LONG_STRINGPARAMETER("pBaseQC", &pBaseQC)
LONG_STRINGPARAMETER("cBaseQC", &cBaseQC)
LONG_PARAMETER_GROUP("Optional Parameters")
LONG_INTPARAMETER("maxNumReads", &maxNumReads)
LONG_PARAMETER("unmapped", &unmapped)
LONG_STRINGPARAMETER("bamIndex", &indexFile)
LONG_STRINGPARAMETER("regionList", ®ionList)
LONG_INTPARAMETER("excludeFlags", &excludeFlags)
LONG_INTPARAMETER("requiredFlags", &requiredFlags)
LONG_PARAMETER("noeof", &noeof)
LONG_PARAMETER("params", ¶ms)
LONG_PARAMETER_GROUP("Optional phred/qual Only Parameters")
LONG_PARAMETER("withinRegion", &withinRegion)
LONG_PARAMETER_GROUP("Optional BaseQC Only Parameters")
LONG_PARAMETER("baseSum", &baseSum)
LONG_INTPARAMETER("bufferSize", &bufferSize)
LONG_INTPARAMETER("minMapQual", &minMapQual)
LONG_STRINGPARAMETER("dbsnp", &dbsnp)
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();
inputParameters.Status();
// In file was not specified but it is mandatory.
std::cerr << "--in is a mandatory argument for stats, "
<< "but was not specified" << std::endl;
return(-1);
}
// Use the index file if unmapped or regionList is not empty.
bool useIndex = (unmapped|| (!regionList.IsEmpty()));
// IndexFile is required, so check to see if it has been set.
if(useIndex && (indexFile == ""))
{
// In file was not specified, so set it to the in file
// + ".bai"
indexFile = inFile + ".bai";
}
////////////////////////////////////////
// Setup in case pileup is used.
Pileup<PileupElementBaseQCStats> pileup(bufferSize);
// Initialize start/end positions.
myStartPos = 0;
myEndPos = -1;
// Open the output qc file if applicable.
IFILE baseQCPtr = NULL;
if(!pBaseQC.IsEmpty() && !cBaseQC.IsEmpty())
{
usage();
inputParameters.Status();
// Cannot specify both types of baseQC.
std::cerr << "Cannot specify both --pBaseQC & --cBaseQC." << std::endl;
return(-1);
}
else if(!pBaseQC.IsEmpty())
{
baseQCPtr = ifopen(pBaseQC, "w");
PileupElementBaseQCStats::setPercentStats(true);
}
else if(!cBaseQC.IsEmpty())
{
baseQCPtr = ifopen(cBaseQC, "w");
PileupElementBaseQCStats::setPercentStats(false);
}
if(baseQCPtr != NULL)
{
PileupElementBaseQCStats::setOutputFile(baseQCPtr);
PileupElementBaseQCStats::printHeader();
}
if((baseQCPtr != NULL) || baseSum)
{
PileupElementBaseQCStats::setMapQualFilter(minMapQual);
PileupElementBaseQCStats::setBaseSum(baseSum);
}
if(params)
{
inputParameters.Status();
}
// Open the file for reading.
SamFile samIn;
if(!samIn.OpenForRead(inFile))
{
fprintf(stderr, "%s\n", samIn.GetStatusMessage());
return(samIn.GetStatus());
}
samIn.SetReadFlags(requiredFlags, excludeFlags);
// Set whether or not basic statistics should be generated.
samIn.GenerateStatistics(basic);
// Read the sam header.
SamFileHeader samHeader;
if(!samIn.ReadHeader(samHeader))
{
fprintf(stderr, "%s\n", samIn.GetStatusMessage());
return(samIn.GetStatus());
}
// Open the bam index file for reading if we are
// doing unmapped reads (also set the read section).
if(useIndex)
{
samIn.ReadBamIndex(indexFile);
if(unmapped)
{
samIn.SetReadSection(-1);
}
if(!regionList.IsEmpty())
{
myRegionList = ifopen(regionList, "r");
}
}
//////////////////////////
// Read dbsnp if specified and doing baseQC
if(((baseQCPtr != NULL) || baseSum) && (!dbsnp.IsEmpty()))
{
// Read the dbsnp file.
IFILE fdbSnp;
fdbSnp = ifopen(dbsnp,"r");
// Determine how many entries.
const SamReferenceInfo& refInfo = samHeader.getReferenceInfo();
int maxRefLen = 0;
for(int i = 0; i < refInfo.getNumEntries(); i++)
{
int refLen = refInfo.getReferenceLength(i);
if(refLen >= maxRefLen)
{
maxRefLen = refLen + 1;
}
}
dbsnpListPtr = new PosList(refInfo.getNumEntries(),maxRefLen);
if(fdbSnp==NULL)
{
std::cerr << "Open dbSNP file " << dbsnp.c_str() << " failed!\n";
}
else if(dbsnpListPtr == NULL)
{
std::cerr << "Failed to init the memory allocation for the dbsnpList.\n";
}
else
{
// Read the dbsnp file.
StringArray tokens;
String buffer;
int position = 0;
int refID = 0;
// Loop til the end of the file.
while (!ifeof(fdbSnp))
{
// Read the next line.
buffer.ReadLine(fdbSnp);
// If it does not have at least 2 columns,
// continue to the next line.
if (buffer.IsEmpty() || buffer[0] == '#') continue;
tokens.AddTokens(buffer);
if(tokens.Length() < 2) continue;
if(!tokens[1].AsInteger(position))
{
std::cerr << "Improperly formatted region line, start position "
<< "(2nd column) is not an integer: "
<< tokens[1]
<< "; Skipping to the next line.\n";
continue;
}
// Look up the reference name.
refID = samHeader.getReferenceID(tokens[0]);
if(refID != SamReferenceInfo::NO_REF_ID)
{
// Reference id was found, so add it to the dbsnp
dbsnpListPtr->addPosition(refID, position);
}
tokens.Clear();
buffer.Clear();
}
}
ifclose(fdbSnp);
}
// Read the sam records.
SamRecord samRecord;
int numReads = 0;
//////////////////////
// Setup in case doing a quality count.
// Quality histogram.
const int MAX_QUAL = 126;
const int START_QUAL = 33;
uint64_t qualCount[MAX_QUAL+1];
for(int i = 0; i <= MAX_QUAL; i++)
{
qualCount[i] = 0;
}
const int START_PHRED = 0;
const int PHRED_DIFF = START_QUAL - START_PHRED;
const int MAX_PHRED = MAX_QUAL - PHRED_DIFF;
uint64_t phredCount[MAX_PHRED+1];
for(int i = 0; i <= MAX_PHRED; i++)
{
phredCount[i] = 0;
}
int refPos = 0;
Cigar* cigarPtr = NULL;
char cigarChar = '?';
// Exclude clips from the qual/phred counts if unmapped reads are excluded.
bool qualExcludeClips = excludeFlags & SamFlag::UNMAPPED;
//////////////////////////////////
// When not reading by sections, getNextSection returns true
// the first time, then false the next time.
while(getNextSection(samIn))
{
// Keep reading records from the file until SamFile::ReadRecord
// indicates to stop (returns false).
while(((maxNumReads < 0) || (numReads < maxNumReads)) && samIn.ReadRecord(samHeader, samRecord))
{
// Another record was read, so increment the number of reads.
++numReads;
// See if the quality histogram should be genereated.
if(qual || phred)
{
// Get the quality.
const char* qual = samRecord.getQuality();
// Check for no quality ('*').
if((qual[0] == '*') && (qual[1] == 0))
{
// This record does not have a quality string, so no
// quality processing is necessary.
}
else
{
int index = 0;
cigarPtr = samRecord.getCigarInfo();
cigarChar = '?';
refPos = samRecord.get0BasedPosition();
if(!qualExcludeClips && (cigarPtr != NULL))
{
// Offset the reference position by any soft clips
// by subtracting the queryIndex of this start position.
// refPos is now the start position of the clips.
refPos -= cigarPtr->getQueryIndex(0);
}
while(qual[index] != 0)
{
// Skip this quality if it is clipped and we are skipping clips.
if(cigarPtr != NULL)
{
cigarChar = cigarPtr->getCigarCharOpFromQueryIndex(index);
}
if(qualExcludeClips && Cigar::isClip(cigarChar))
{
// Skip a clipped quality.
++index;
// Increment the position.
continue;
}
if(withinRegion && (myEndPos != -1) && (refPos >= myEndPos))
{
// We have hit the end of the region, stop processing this
// quality string.
break;
}
if(withinRegion && (refPos < myStartPos))
{
// This position is not in the target.
++index;
// Update the position if this is found in the reference or a clip.
if(Cigar::foundInReference(cigarChar) || Cigar::isClip(cigarChar))
{
++refPos;
}
continue;
}
// Check for valid quality.
if((qual[index] < START_QUAL) || (qual[index] > MAX_QUAL))
{
if(qual)
{
std::cerr << "Invalid Quality found: " << qual[index]
<< ". Must be between "
<< START_QUAL << " and " << MAX_QUAL << ".\n";
}
if(phred)
{
std::cerr << "Invalid Phred Quality found: " << qual[index] - PHRED_DIFF
<< ". Must be between "
<< START_QUAL << " and " << MAX_QUAL << ".\n";
}
// Skip an invalid quality.
++index;
// Update the position if this is found in the reference or a clip.
if(Cigar::foundInReference(cigarChar) || Cigar::isClip(cigarChar))
{
++refPos;
}
continue;
}
// Increment the count for this quality.
++(qualCount[(int)(qual[index])]);
++(phredCount[(int)(qual[index]) - PHRED_DIFF]);
// Update the position if this is found in the reference or a clip.
if(Cigar::foundInReference(cigarChar) || Cigar::isClip(cigarChar))
{
++refPos;
}
++index;
}
}
}
// Check the next thing to do for the read.
if((baseQCPtr != NULL) || baseSum)
{
// Pileup the bases for this read.
pileup.processAlignmentRegion(samRecord, myStartPos, myEndPos, dbsnpListPtr);
}
}
// Done with a section, move on to the next one.
// New section, so flush the pileup.
pileup.flushPileup();
}
// Flush the rest of the pileup.
if((baseQCPtr != NULL) || baseSum)
{
// Pileup the bases.
pileup.processAlignmentRegion(samRecord, myStartPos, myEndPos, dbsnpListPtr);
PileupElementBaseQCStats::printSummary();
ifclose(baseQCPtr);
}
std::cerr << "Number of records read = " <<
samIn.GetCurrentRecordCount() << std::endl;
if(basic)
{
std::cerr << std::endl;
samIn.PrintStatistics();
}
// Print the quality stats.
if(qual)
{
std::cerr << std::endl;
std::cerr << "Quality\tCount\n";
for(int i = START_QUAL; i <= MAX_QUAL; i++)
{
std::cerr << i << "\t" << qualCount[i] << std::endl;
}
}
// Print the phred quality stats.
if(phred)
{
std::cerr << std::endl;
std::cerr << "Phred\tCount\n";
for(int i = START_PHRED; i <= MAX_PHRED; i++)
{
std::cerr << i << "\t" << phredCount[i] << std::endl;
}
}
SamStatus::Status status = samIn.GetStatus();
if(status == SamStatus::NO_MORE_RECS)
{
// A status of NO_MORE_RECS means that all reads were successful.
status = SamStatus::SUCCESS;
}
return(status);
}
// 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 GapInfo::processFile(const char* inputFileName, const char* outputFileName,
const char* refFile, bool detailed,
bool checkFirst, bool checkStrand)
{
// Open the file for reading.
SamFile samIn;
samIn.OpenForRead(inputFileName);
// Read the sam header.
SamFileHeader samHeader;
samIn.ReadHeader(samHeader);
SamRecord samRecord;
GenomeSequence* refPtr = NULL;
if(strcmp(refFile, "") != 0)
{
refPtr = new GenomeSequence(refFile);
}
IFILE outFile = ifopen(outputFileName, "w");
// Map for summary.
std::map<int, int> gapInfoMap;
// Keep reading records until ReadRecord returns false.
while(samIn.ReadRecord(samHeader, samRecord))
{
uint16_t samFlags = samRecord.getFlag();
if((!SamFlag::isMapped(samFlags)) ||
(!SamFlag::isMateMapped(samFlags)) ||
(!SamFlag::isPaired(samFlags)) ||
(samFlags & SamFlag::SECONDARY_ALIGNMENT) ||
(SamFlag::isDuplicate(samFlags)) ||
(SamFlag::isQCFailure(samFlags)))
{
// unmapped, mate unmapped, not paired,
// not the primary alignment,
// duplicate, fails vendor quality check
continue;
}
// No gap info if the chromosome names are different or
// are unknown.
int32_t refID = samRecord.getReferenceID();
if((refID != samRecord.getMateReferenceID()) || (refID == -1))
{
continue;
}
int32_t readStart = samRecord.get0BasedPosition();
int32_t mateStart = samRecord.get0BasedMatePosition();
// If the mate starts first, then the pair was processed by
// the mate.
if(mateStart < readStart)
{
continue;
}
if((mateStart == readStart) && (SamFlag::isReverse(samFlags)))
{
// read and mate start at the same position, so
// only process the forward strand.
continue;
}
// Process this read pair.
int32_t readEnd = samRecord.get0BasedAlignmentEnd();
int32_t gapSize = mateStart - readEnd - 1;
if(detailed)
{
// Output the gap info.
ifprintf(outFile, "%s\t%d\t%d",
samRecord.getReferenceName(), readEnd+1, gapSize);
// Check if it is not the first or if it is not the forward strand.
if(checkFirst && !SamFlag::isFirstFragment(samFlags))
{
ifprintf(outFile, "\tNotFirst");
}
if(checkStrand && SamFlag::isReverse(samFlags))
{
ifprintf(outFile, "\tReverse");
}
ifprintf(outFile, "\n");
}
else
{
// Summary.
// Skip reads that are not the forward strand.
if(SamFlag::isReverse(samFlags))
{
// continue
continue;
}
// Forward.
// Check the reference for 'N's.
if(refPtr != NULL)
{
genomeIndex_t chromStartIndex =
refPtr->getGenomePosition(samRecord.getReferenceName());
if(chromStartIndex == INVALID_GENOME_INDEX)
{
// Invalid position, so continue to the next one.
continue;
}
bool skipRead = false;
for(int i = readEnd + 1; i < mateStart; i++)
{
if((*refPtr)[i] == 'N')
{
// 'N' in the reference, so continue to the next read.
skipRead = true;
break;
}
}
if(skipRead)
{
continue;
}
}
// Update the gapInfo.
gapInfoMap[gapSize]++;
}
}
if(!detailed)
{
// Output the summary.
ifprintf(outFile, "GapSize\tNumPairs\n");
for(std::map<int,int>::iterator iter = gapInfoMap.begin();
iter != gapInfoMap.end(); iter++)
{
ifprintf(outFile, "%d\t%d\n", (*iter).first, (*iter).second);
}
}
SamStatus::Status returnStatus = samIn.GetStatus();
if(returnStatus == SamStatus::NO_MORE_RECS)
{
return(SamStatus::SUCCESS);
}
return(returnStatus);
}
