bool GetNextAlignment(BamAlignment& al, BamMultiReader& reader, int32_t refID)
{
bool good = reader.GetNextAlignmentCore(al);
if(not good or al.RefID != refID)
return false;
else if(not al.IsMapped() or al.IsDuplicate() or al.IsFailedQC())
return GetNextAlignment(al, reader, refID);
else
return true;
}
// merges sorted temp BAM files into single sorted output BAM file
bool SortTool::SortToolPrivate::MergeSortedRuns(void) {
// open up multi reader for all of our temp files
// this might get broken up if we do a multi-pass system later ??
BamMultiReader multiReader;
if ( !multiReader.Open(m_tempFilenames) ) {
cerr << "bamtools sort ERROR: could not open BamMultiReader for merging temp files... Aborting."
<< endl;
return false;
}
// set sort order for merge
if ( m_settings->IsSortingByName )
multiReader.SetSortOrder(BamMultiReader::SortedByReadName);
else
multiReader.SetSortOrder(BamMultiReader::SortedByPosition);
// open writer for our completely sorted output BAM file
BamWriter mergedWriter;
if ( !mergedWriter.Open(m_settings->OutputBamFilename, m_headerText, m_references) ) {
cerr << "bamtools sort ERROR: could not open " << m_settings->OutputBamFilename
<< " for writing... Aborting." << endl;
multiReader.Close();
return false;
}
// while data available in temp files
BamAlignment al;
while ( multiReader.GetNextAlignmentCore(al) )
mergedWriter.SaveAlignment(al);
// close readers
multiReader.Close();
mergedWriter.Close();
// delete all temp files
vector<string>::const_iterator tempIter = m_tempFilenames.begin();
vector<string>::const_iterator tempEnd = m_tempFilenames.end();
for ( ; tempIter != tempEnd; ++tempIter ) {
const string& tempFilename = (*tempIter);
remove(tempFilename.c_str());
}
return true;
}
bool StatsTool::StatsToolPrivate::Run() {
// opens the BAM files without checking for indexes
BamMultiReader reader;
if ( !reader.Open(settings->InputFiles, false, true) ) {
cerr << "Could not open input BAM file(s)... quitting." << endl;
reader.Close();
return false;
}
// plow through file, keeping track of stats
BamAlignment al;
while ( reader.GetNextAlignmentCore(al) )
ProcessAlignment(al);
// print stats
PrintStats();
// clean and exit
reader.Close();
return true;
}
bool RandomTool::RandomToolPrivate::Run(void) {
// set to default stdin if no input files provided
if ( !m_settings->HasInput && !m_settings->HasInputFilelist )
m_settings->InputFiles.push_back(Options::StandardIn());
// add files in the filelist to the input file list
if ( m_settings->HasInputFilelist ) {
ifstream filelist(m_settings->InputFilelist.c_str(), ios::in);
if ( !filelist.is_open() ) {
cerr << "bamtools random ERROR: could not open input BAM file list... Aborting." << endl;
return false;
}
string line;
while ( getline(filelist, line) )
m_settings->InputFiles.push_back(line);
}
// open our reader
BamMultiReader reader;
if ( !reader.Open(m_settings->InputFiles) ) {
cerr << "bamtools random ERROR: could not open input BAM file(s)... Aborting." << endl;
return false;
}
// look up index files for all BAM files
reader.LocateIndexes();
// make sure index data is available
if ( !reader.HasIndexes() ) {
cerr << "bamtools random ERROR: could not load index data for all input BAM file(s)... Aborting." << endl;
reader.Close();
return false;
}
// get BamReader metadata
const string headerText = reader.GetHeaderText();
const RefVector references = reader.GetReferenceData();
if ( references.empty() ) {
cerr << "bamtools random ERROR: no reference data available... Aborting." << endl;
reader.Close();
return false;
}
// determine compression mode for BamWriter
bool writeUncompressed = ( m_settings->OutputFilename == Options::StandardOut() &&
!m_settings->IsForceCompression );
BamWriter::CompressionMode compressionMode = BamWriter::Compressed;
if ( writeUncompressed ) compressionMode = BamWriter::Uncompressed;
// open BamWriter
BamWriter writer;
writer.SetCompressionMode(compressionMode);
if ( !writer.Open(m_settings->OutputFilename, headerText, references) ) {
cerr << "bamtools random ERROR: could not open " << m_settings->OutputFilename
<< " for writing... Aborting." << endl;
reader.Close();
return false;
}
// if user specified a REGION constraint, attempt to parse REGION string
BamRegion region;
if ( m_settings->HasRegion && !Utilities::ParseRegionString(m_settings->Region, reader, region) ) {
cerr << "bamtools random ERROR: could not parse REGION: " << m_settings->Region << endl;
cerr << "Check that REGION is in valid format (see documentation) and that the coordinates are valid"
<< endl;
reader.Close();
writer.Close();
return false;
}
// seed our random number generator
srand( time(NULL) );
// grab random alignments
BamAlignment al;
unsigned int i = 0;
while ( i < m_settings->AlignmentCount ) {
int randomRefId = 0;
int randomPosition = 0;
// use REGION constraints to select random refId & position
if ( m_settings->HasRegion ) {
// select a random refId
randomRefId = getRandomInt(region.LeftRefID, region.RightRefID);
// select a random position based on randomRefId
const int lowerBoundPosition = ( (randomRefId == region.LeftRefID)
? region.LeftPosition
: 0 );
const int upperBoundPosition = ( (randomRefId == region.RightRefID)
? region.RightPosition
: (references.at(randomRefId).RefLength - 1) );
randomPosition = getRandomInt(lowerBoundPosition, upperBoundPosition);
}
// otherwise select from all possible random refId & position
else {
// select random refId
randomRefId = getRandomInt(0, (int)references.size() - 1);
// select random position based on randomRefId
const int lowerBoundPosition = 0;
const int upperBoundPosition = references.at(randomRefId).RefLength - 1;
randomPosition = getRandomInt(lowerBoundPosition, upperBoundPosition);
}
// if jump & read successful, save first alignment that overlaps random refId & position
if ( reader.Jump(randomRefId, randomPosition) ) {
while ( reader.GetNextAlignmentCore(al) ) {
if ( al.RefID == randomRefId && al.Position >= randomPosition ) {
writer.SaveAlignment(al);
++i;
break;
}
}
}
}
// cleanup & exit
reader.Close();
writer.Close();
return true;
}
int CountTool::Run(int argc, char* argv[]) {
// parse command line arguments
Options::Parse(argc, argv, 1);
// if no '-in' args supplied, default to stdin
if ( !m_settings->HasInput )
m_settings->InputFiles.push_back(Options::StandardIn());
// open reader without index
BamMultiReader reader;
if (!reader.Open(m_settings->InputFiles, false, true)) {
cerr << "ERROR: Could not open input BAM file(s)... Aborting." << endl;
return 1;
}
// alignment counter
BamAlignment al;
int alignmentCount(0);
// if no region specified, count entire file
if ( !m_settings->HasRegion ) {
while ( reader.GetNextAlignmentCore(al) )
++alignmentCount;
}
// otherwise attempt to use region as constraint
else {
// if region string parses OK
BamRegion region;
if ( Utilities::ParseRegionString(m_settings->Region, reader, region) ) {
// attempt to re-open reader with index files
reader.Close();
bool openedOK = reader.Open(m_settings->InputFiles, true, true );
// if error
if ( !openedOK ) {
cerr << "ERROR: Could not open input BAM file(s)... Aborting." << endl;
return 1;
}
// if index data available, we can use SetRegion
if ( reader.IsIndexLoaded() ) {
// attempt to use SetRegion(), if failed report error
if ( !reader.SetRegion(region.LeftRefID, region.LeftPosition, region.RightRefID, region.RightPosition) ) {
cerr << "ERROR: Region requested, but could not set BamReader region to REGION: " << m_settings->Region << " Aborting." << endl;
reader.Close();
return 1;
}
// everything checks out, just iterate through specified region, counting alignments
while ( reader.GetNextAlignmentCore(al) )
++alignmentCount;
}
// no index data available, we have to iterate through until we
// find overlapping alignments
else {
while( reader.GetNextAlignmentCore(al) ) {
if ( (al.RefID >= region.LeftRefID) && ( (al.Position + al.Length) >= region.LeftPosition ) &&
(al.RefID <= region.RightRefID) && ( al.Position <= region.RightPosition) )
{
++alignmentCount;
}
}
}
}
// error parsing REGION string
else {
cerr << "ERROR: Could not parse REGION - " << m_settings->Region << endl;
cerr << "Be sure REGION is in valid format (see README) and that coordinates are valid for selected references" << endl;
reader.Close();
return 1;
}
}
// print results
cout << alignmentCount << endl;
// clean & exit
reader.Close();
return 0;
}
int bbctools_create( BbcUtils::OptParser &optParser ) {
const vector<string> cmdArgs = optParser.getArgs();
// remove .bbc extension from bbc file root, if present
string bbcfileRoot = optParser.getOptValue( "bbc" );
int i = bbcfileRoot.size() - 4;
if( i > 0 && bbcfileRoot.substr(i,4) == ".bbc" ) {
bbcfileRoot = bbcfileRoot.substr(0,i);
}
bool f_bci = optParser.getOptBoolean("index");
bool f_cbc = optParser.getOptBoolean("coarse");
string targetRegions = optParser.getOptValue("regions");
string annotationFields = optParser.getOptValue( "annotationFields");
vector<string> auxRegionSplit = BbcUtils::mapToPairList(annotationFields);
string sumstatsFile = optParser.getOptValue("sumStats");
string covstatsFile = optParser.getOptValue("covStats");
string readOrigFile = optParser.getOptValue("readOrigin");
string readType = optParser.getOptValue("readType");
string covDepths = optParser.getOptValue("covDepths","-");
double minPcCov = optParser.getOptNumber("minPcCov");
int32_t primerLength = optParser.getOptInteger( "primerLength", (readType == "AmpliSeq" ? 30 : 0) );
int32_t maxE2eEndGap = optParser.getOptInteger( "e2eGap", (readType == "AmpliSeq" ? 2 : 0) );
bool autoCreateBamIndex = optParser.getOptBoolean("autoCreateBamIndex");
bool samdepth = optParser.getOptBoolean("samdepth");
int32_t filterQuality = optParser.getOptInteger("minMAPQ");
int32_t minAlignLength = optParser.getOptInteger("minAlignLength");
bool filterDuplicates = optParser.getOptBoolean("noDups");
bool filterUnique = optParser.getOptBoolean("unique");
uint32_t skipFlag = filterDuplicates ? 0x704 : 0x304;
uint16_t minMapQuality = filterUnique ? 1 : filterQuality;
bool onlyOnTargetReads = optParser.getOptBoolean("onTargetReads");
bool onlyOnTargetBases = optParser.getOptBoolean("onTargetBases");
// possible future options
bool invertOnTarget = false;
// check basic valid argument values and combinations
int numOuts = !bbcfileRoot.empty() + !covstatsFile.empty() + !sumstatsFile.empty() + !readOrigFile.empty();
int numPipes = (bbcfileRoot == "-") + (covstatsFile == "-") + (sumstatsFile == "-") + (readOrigFile == "-");
if( numOuts == 0 && !f_bci && !f_cbc ) {
bbcfileRoot = "-"; // default if no other output specified
} else if( numPipes > 1 ) {
cerr << "Error: bbctools create: Only one file output (--covStats, --sumStats, --readOrigin or --bbc) may be piped to STDOUT." << endl;
return -1;
} else if( samdepth && numOuts ) {
cerr << "Error: bbctools create: --samdepth (-s) option may only be used without other output options." << endl;
return -1;
}
// check if single argument is a BBC file and leave open for reading if so
BbcView bbcView;
bool haveBbcFile = cmdArgs.size() == 1 && bbcView.Open( cmdArgs[0], true );
bbcView.SelectPrintStream( samdepth ? "SAMDEPTH" : "BBCVIEW" );
// check distinction between default and explicit no target regions - only for BBC input
bool explicitNoTargetRegions = false;
if( targetRegions == "-" ) {
explicitNoTargetRegions = haveBbcFile;
targetRegions = "";
}
if( targetRegions.empty() ) {
if( onlyOnTargetBases && explicitNoTargetRegions && !invertOnTarget ) {
cerr << "Warning: bbctools create --onTargetBases (-b) option with --regions '-' produces no coverage." << endl;
} else if( onlyOnTargetReads ) {
cerr << "Error: bbctools create --onTargetReads (-r) option requires a --regions file." << endl;
return -1;
}
}
// check for legal BBC create options
if( f_bci || f_cbc ) {
if( (bbcfileRoot.empty() || bbcfileRoot == "-") && !haveBbcFile ) {
string opt = f_bci ? "--index (-i)" : "--coarse (-c)";
cerr << "Error: bbctools create "+opt+" option requires the --bbc (-B) option or a BBC source file." << endl;
return -1;
}
}
BamMultiReader bamReader;
if( haveBbcFile ) {
// warn for options that do not work with BBC input
if( filterQuality > 0 || filterDuplicates || filterUnique || minAlignLength ) {
cerr << "Warning: SAM flag, alignment length and MAPQ filters ignored for BBC source file." << endl;
}
if( samdepth ) {
cerr << "Error: --samdepth option is not supported for BBC source files." << endl;
return -1;
}
if( !readOrigFile.empty() ) {
cerr << "Error: --readOrigin option is not supported for BBC source files." << endl;
return -1;
}
} else {
// check / open for multiple BAM file inputs
if ( !bamReader.Open(cmdArgs) ) {
if( cmdArgs.size() == 1 ) cerr << "ERROR: Could not read input BAM file:";
else cerr << "ERROR: Could not read all input BAM files:";
// get and clean up bamtools error msg
string errMsg = bamReader.GetErrorString();
size_t i = errMsg.find_first_of('\n');
if( i != string::npos ) errMsg = errMsg.substr(i+1);
i = errMsg.find("::");
if( i != string::npos ) {
i = errMsg.find(": ");
if( i != string::npos ) errMsg = errMsg.substr(i+1);
}
errMsg = BbcUtils::stringTrim(errMsg);
errMsg[0] = toupper(errMsg[0]);
cerr << endl << errMsg << "." << endl;
return 1;
}
}
// grab reference list from either input source
const RefVector &references = haveBbcFile ? bbcView.GetReferenceData() : bamReader.GetReferenceData();
if( !references.size() ) {
// Issue would already been detected if input was BBC file
cerr << "ERROR: " << (cmdArgs.size() > 1 ? "One or more " : "");
cerr << "BAM file contains unaligned reads (no references).\n";
return 1;
}
// check/set up target regions input regions/region statistics output
RegionCoverage *regions = NULL;
string covstatsStaticFields;
bool trackRegionBaseCov = !covDepths.empty();
if( covstatsFile.empty() ) {
trackRegionBaseCov = false;
if( !annotationFields.empty() ) {
cerr << "Warning: --annotationFields (A) option ignored without --covStats (-C) option." << endl;
}
if( !covDepths.empty() && covDepths != "-" ) {
cerr << "Warning: --covDepths (-D) option ignored without --covStats (-C) option." << endl;
}
if( !readType.empty() ) {
cerr << "Warning: --readType (-T) option ignored without --covStats (-C) option." << endl;
}
// read regions for input only and/or creating sumStats
if( !targetRegions.empty() || explicitNoTargetRegions || !sumstatsFile.empty() ) {
regions = new RegionCoverage(references);
}
} else if( readType == "trgreads" || readType == "amplicon" || readType == "AmpliSeq" ) {
if( haveBbcFile ) {
cerr << "Creation of read coverage requires BAM file input." << endl;
return -1;
}
AmpliconRegionStatistics *ampRegionStats = new AmpliconRegionStatistics(references);
ampRegionStats->SetGenericReads( readType == "trgreads" );
ampRegionStats->SetSigFacCoverage( minPcCov/100 );
ampRegionStats->SetMaxUpstreamPrimerStart( primerLength );
ampRegionStats->SetMaxE2eEndDistance( maxE2eEndGap );
covstatsStaticFields = "overlaps,";
covstatsStaticFields += (minPcCov > 0) ? "fwd_cov,rev_cov" : "fwd_e2e,rev_e2e";
covstatsStaticFields += ",total_reads,fwd_reads,rev_reads";
regions = ampRegionStats;
} else if( readType == "trgbases" ) {
if( haveBbcFile && targetRegions.empty() && !explicitNoTargetRegions ) {
cerr << "Warning: Assuming reference contigs for base coverage targets (=> option --regions -)" << endl;
}
RegionStatistics *regionStats = new RegionStatistics(references);
covstatsStaticFields = "covered,uncov_5p,uncov_3p,ave_basereads,fwd_basereads,rev_basereads";
trackRegionBaseCov = true;
regions = regionStats;
} else if( readType == "covdepth" || readType.empty() ) {
// output (sorted) targets file with only covDepth stats (if any)
regions = new RegionCoverage(references);
} else {
cerr << "Unknown read type '" << readType << "'" << endl;
return -1;
}
// Load the input regions or default to whole reference contig targets
if( regions ) {
regions->SetCovAtDepths( covDepths == "-" ? "20,100,500" : covDepths );
if( targetRegions.empty() ) {
regions->SetWholeContigTargets();
// set contigs as explicit regions means all reads will seen as on-target
// for consistency these are inverted (for input from BBC)
invertOnTarget = true;
} else {
string auxFieldIdx = auxRegionSplit.size() ? auxRegionSplit[0] : "";
string errMsg = regions->Load( targetRegions, "BED", auxFieldIdx );
if( !errMsg.empty() ) {
cerr << "ERROR: " + errMsg + "\n";
return 1;
}
}
if( onlyOnTargetReads && haveBbcFile ) {
cerr << "Error: bbctools create --onTargetReads option is not supported for BBC source file." << endl;
return -1;
}
}
//
// Perform all bbctools create utilities
//
BbcCreate *bbcCreate = NULL;
if( !bbcfileRoot.empty() && (bbcfileRoot != "-" || !haveBbcFile) ) {
bbcCreate = new BbcCreate(references);
if( bbcfileRoot != "-" && !bbcCreate->Open(bbcfileRoot+".bbc") ) {
return 1;
}
bbcCreate->SetNoOffTargetPositions(onlyOnTargetBases);
}
bbcView.SetNoOffTargetPositions(onlyOnTargetBases);
// Stream input to output creators
if( haveBbcFile ) {
// BBC reader and driver via BbcView object
if( bbcfileRoot != "-" || !covstatsFile.empty() ) {
// disable BbcView text stream if using for file creation
bbcView.SelectPrintStream("NONE");
}
// process input BBC for just new BBC and target coverage (defer BCI/CBC)
bbcView.SetBbcCreate(bbcCreate);
bbcView.SetRegionCoverage(regions);
// explicitNoTargetRegions intended for explicitly removing on-target coverage
bbcView.SetInvertOnTarget(explicitNoTargetRegions ^ invertOnTarget);
if( bbcCreate || regions || bbcfileRoot == "-" ) {
bbcView.ReadAll();
}
} else {
// Test read tracking option for file write
TrackReads *readTracker = NULL;
try {
if( !readOrigFile.empty() )
readTracker = new TrackReads( readOrigFile, regions );
} catch( std::runtime_error & ) {
cerr << "ERROR: Unable to write to read tracking file " << readOrigFile << endl;
return 1;
}
// BAM reader, BaseCoverage driver, dispatching to BbcCreate and BbcView objects
BaseCoverage baseCov(references);
baseCov.SetRegionCoverage(regions);
baseCov.SetBbcCreate(bbcCreate);
baseCov.SetInvertOnTarget(invertOnTarget);
if( bbcfileRoot == "-" ) {
baseCov.SetBbcView(&bbcView);
}
// Certain options require that all reads are processed, invalidating other performance options
bool trackAllReads = !sumstatsFile.empty() || readTracker;
// Implicit set of onlyOnTargetReads for performance when only these reads are required
bool useBaseCov = (bbcfileRoot == "-" || bbcCreate);
if( !targetRegions.empty() && !trackAllReads ) {
onlyOnTargetReads |= onlyOnTargetBases;
if( samdepth || !useBaseCov ) onlyOnTargetReads = true;
}
useBaseCov |= trackRegionBaseCov;
// do not allow jumping if sumStats option is used - need to count all reads
bool bamReaderSetRegions = (s_useBamReaderJump && !trackAllReads);
int trgContig = 0, trgSrtPos = 0, trgEndPos = 0;
int minJumpLen = s_initialMinJumpLen;
int maxReadLen = s_initialMaxReadLen;
if( onlyOnTargetReads ) {
// load/create BAM index files for targeted reading
// Note: BamIndex::BAMTOOLS format performed very badly and cannot use mixed with BTI/BAI files
if( bamReaderSetRegions && !bamReader.LocateIndexes() ) {
string plural( cmdArgs.size() > 1 ? "s" : "" );
if( autoCreateBamIndex ) {
cerr << "Warning: Did not locate BAM index (BAI) file" << plural << ", creating bamtools version..." << endl;
// to avoid bug use new instance of BamMultiReader
BamMultiReader bamReader2;
if( !bamReader2.Open(cmdArgs) || !bamReader2.CreateIndexes() ) {
cerr << "WARNING: Failed to create BAM index file" << plural << "." << endl;
bamReaderSetRegions = false;
} else {
if( cmdArgs.size() == 1 ) {
cerr << "Successfully created BAM index file: " << BbcUtils::fileName(cmdArgs[0]) << ".bai" << endl;
} else {
cerr << "Successfully created BAM index files." << endl;
}
// re-locate indexes with first reader - could not seem to locate BTI files created!
if( !bamReader.LocateIndexes() ) {
cerr << "WARNING: Failed to locate BAM index file" << plural << " just created!" << endl;
bamReaderSetRegions = false;
}
}
} else {
cerr << "Warning: BAM index file" << plural << " not located for targeted BAM access." << endl;
bamReaderSetRegions = false;
}
}
// cancel region filtering if there are no regions to iterate (unexpected)
if( !regions->GetNextRegion( trgContig, trgSrtPos, trgEndPos ) ) {
onlyOnTargetReads = bamReaderSetRegions = false;
}
if( bamReaderSetRegions ) {
bamReader.Jump( trgContig, trgSrtPos-maxReadLen );
}
}
BamAlignment aln;
while( bamReader.GetNextAlignmentCore(aln) ) {
// appears to be an undocumented behavior here
if( aln.RefID < 0 ) continue;
// skip filtered reads by flag, length or mapping quality
if( aln.AlignmentFlag & skipFlag ) continue;
if( aln.MapQuality < minMapQuality ) continue;
int32_t endPos = aln.GetEndPosition();
if( minAlignLength > 0 ) {
if( endPos - aln.Position < minAlignLength ) continue;
}
// screen for on-target reads
if( onlyOnTargetReads ) {
// find next region overlapping or beyond of current read
bool moreRegions = true;
bool setRegion = false;
while( aln.RefID > trgContig || (aln.RefID == trgContig && aln.Position > trgEndPos) ) {
if( !regions->GetNextRegion( trgContig, trgSrtPos, trgEndPos ) ) {
moreRegions = false;
break;
}
setRegion = bamReaderSetRegions;
}
if( !moreRegions ) {
// prevent further on-target checks and exit early if not using sumStats
onlyOnTargetReads = false;
if( trackAllReads ) {
// force tracking of off-target reads
regions->TrackReadsOnRegion(aln,endPos);
if( readTracker ) readTracker->Write(aln,endPos);
continue;
}
break;
}
if( setRegion ) {
// track max read length for future index jumps - just in case long reads ever used
if( endPos - aln.Position > maxReadLen ) {
maxReadLen = endPos - aln.Position;
if( maxReadLen > minJumpLen ) minJumpLen = maxReadLen;
}
if( aln.RefID != trgContig || trgSrtPos - aln.Position > minJumpLen ) {
bamReader.Jump( trgContig, trgSrtPos-maxReadLen );
}
}
if( aln.RefID < trgContig || endPos < trgSrtPos ) {
// force tracking of off-target reads
if( trackAllReads ) {
regions->TrackReadsOnRegion(aln,endPos);
if( readTracker ) readTracker->Write(aln,endPos);
}
continue; // current is before next target region - fetch the next within bounds
}
}
// record base coverage and region coverage statistics
if( useBaseCov ) {
endPos = baseCov.AddAlignment(aln,endPos);
if( endPos <= 0 ) {
if( endPos == 0 ) continue; // read was silently ignored
cerr << "ERROR: BAM file is not correctly sorted vs. reference." << endl;
return 1;
}
}
// record read coverage and region coverage statistics
if( regions ) {
regions->TrackReadsOnRegion(aln,endPos);
}
if( readTracker ) {
readTracker->Write(aln,endPos);
}
}
// flush and close objects associated with output
baseCov.Flush();
}
// Output in-memory region stats file and ensure BBC file is closed
if( regions ) {
// build output fields title string
string outFields = "contig_id,contig_srt,contig_end";
if( !auxRegionSplit.empty() ) outFields += "," + auxRegionSplit[1];
if( !covstatsStaticFields.empty() ) outFields += "," + covstatsStaticFields;
regions->Write( covstatsFile, outFields );
if( !sumstatsFile.empty() ) {
regions->WriteSummary( sumstatsFile, invertOnTarget );
}
delete regions;
}
delete bbcCreate;
// Complete remaining file creation options using a BBC file input
// NOTE: Using BbbCreate for this would require code duplication and concurrent file output streaming
if( f_bci || f_cbc ) {
// Check BBC file source
if( haveBbcFile ) {
bbcfileRoot = cmdArgs[0];
int i = bbcfileRoot.size() - 4;
if( i > 0 && bbcfileRoot.substr(i,4) == ".bbc" ) {
bbcfileRoot = bbcfileRoot.substr(0,i);
}
} else if( !bbcView.Open( bbcfileRoot+".bbc", true ) ) {
cerr << "ERROR: Unexpected failure to read new BBC file '"+bbcfileRoot+".bam'" << endl;
return 1;
}
if( f_bci ) {
BbcIndex indexer( bbcfileRoot+".bci" );
if( !bbcView.CreateIndex(indexer) ) {
cerr << "ERROR: Failed to create index file '" << bbcfileRoot << ".bci'" << endl;
return 1;
}
}
if( f_cbc ) {
// CBC generation can use BCI file but is no faster since whole BBC file is read
BbcCoarse cbcWriter( bbcfileRoot+".cbc" );
if( !bbcView.CreateCbc(cbcWriter) ) {
cerr << "ERROR: Failed to create coarse base coverage file '" << bbcfileRoot << ".cbc'" << endl;
return 1;
}
}
}
return 0;
}
bool MergeTool::MergeToolPrivate::Run(void) {
// set to default input if none provided
if ( !m_settings->HasInputBamFilename )
m_settings->InputFiles.push_back(Options::StandardIn());
// opens the BAM files (by default without checking for indexes)
BamMultiReader reader;
if ( !reader.Open(m_settings->InputFiles) ) {
cerr << "bamtools merge ERROR: could not open input BAM file(s)... Aborting." << endl;
return false;
}
// retrieve header & reference dictionary info
std::string mergedHeader = reader.GetHeaderText();
RefVector references = reader.GetReferenceData();
// determine compression mode for BamWriter
bool writeUncompressed = ( m_settings->OutputFilename == Options::StandardOut() &&
!m_settings->IsForceCompression );
BamWriter::CompressionMode compressionMode = BamWriter::Compressed;
if ( writeUncompressed ) compressionMode = BamWriter::Uncompressed;
// open BamWriter
BamWriter writer;
writer.SetCompressionMode(compressionMode);
if ( !writer.Open(m_settings->OutputFilename, mergedHeader, references) ) {
cerr << "bamtools merge ERROR: could not open "
<< m_settings->OutputFilename << " for writing." << endl;
reader.Close();
return false;
}
// if no region specified, store entire contents of file(s)
if ( !m_settings->HasRegion ) {
BamAlignment al;
while ( reader.GetNextAlignmentCore(al) )
writer.SaveAlignment(al);
}
// otherwise attempt to use region as constraint
else {
// if region string parses OK
BamRegion region;
if ( Utilities::ParseRegionString(m_settings->Region, reader, region) ) {
// attempt to find index files
reader.LocateIndexes();
// if index data available for all BAM files, we can use SetRegion
if ( reader.HasIndexes() ) {
// attempt to use SetRegion(), if failed report error
if ( !reader.SetRegion(region.LeftRefID,
region.LeftPosition,
region.RightRefID,
region.RightPosition) )
{
cerr << "bamtools merge ERROR: set region failed. Check that REGION describes a valid range"
<< endl;
reader.Close();
return false;
}
// everything checks out, just iterate through specified region, storing alignments
BamAlignment al;
while ( reader.GetNextAlignmentCore(al) )
writer.SaveAlignment(al);
}
// no index data available, we have to iterate through until we
// find overlapping alignments
else {
BamAlignment al;
while ( reader.GetNextAlignmentCore(al) ) {
if ( (al.RefID >= region.LeftRefID) && ( (al.Position + al.Length) >= region.LeftPosition ) &&
(al.RefID <= region.RightRefID) && ( al.Position <= region.RightPosition) )
{
writer.SaveAlignment(al);
}
}
}
}
// error parsing REGION string
else {
cerr << "bamtools merge ERROR: could not parse REGION - " << m_settings->Region << endl;
cerr << "Check that REGION is in valid format (see documentation) and that the coordinates are valid"
<< endl;
reader.Close();
writer.Close();
return false;
}
}
// clean & exit
reader.Close();
writer.Close();
return true;
}
int FileReader::runInternal()
{
ogeNameThread("am_FileReader");
if(!format_specified)
format = deduceFileFormat();
if(format == FORMAT_BAM)
{
BamMultiReader reader;
if(!reader.Open(filenames)) {
cerr << "Error opening BAM files." << endl;
reader.Close();
return -1;
}
header = reader.GetHeader();
references = reader.GetReferenceData();
open = true;
BamAlignment * al;
while(true)
{
if(load_string_data)
al = reader.GetNextAlignment();
else
al = reader.GetNextAlignmentCore();
if(!al)
break;
putOutputAlignment(al);
}
reader.Close();
} else if(format == FORMAT_SAM) {
vector<SamReader> readers;
SamHeader first_header;
// before doing any reading, open the files to
// verify they are the right format, etc.
for(int i = 0; i < filenames.size(); i++) {
SamReader reader;
if(!reader.Open(filenames[i])) {
cerr << "Error opening SAM file: " << filenames[i] << endl;
return -1;
}
if(filenames.size() > 1 && i == 0)
first_header = header;
// TODO: We can probably find a better way to deal with multiple SAM file headers,
// but for now we should disallow different headers to avoid issues.
if(i > 0 && header.ToString() != first_header.ToString())
cerr << "Warning! SAM input files have different headers." << endl;
reader.Close();
}
for(int i = 0; i < filenames.size(); i++) {
SamReader reader;
if(!reader.Open(filenames[i])) {
cerr << "Error opening SAM file: " << filenames[i] << endl;
return -1;
}
header = reader.GetHeader();
references = reader.GetReferenceData();
open = true;
if(filenames.size() > 1 && i == 0)
first_header = header;
BamAlignment * al = NULL;
while(true)
{
al = reader.GetNextAlignment();
if(NULL == al)
break;
putOutputAlignment(al);
}
reader.Close();
}
} else {
cerr << "FileReader couldn't detect file format. Aborting." << endl;
exit(-1);
return -1;
}
return 0;
}
int MergeTool::Run(int argc, char* argv[]) {
// parse command line arguments
Options::Parse(argc, argv, 1);
// set to default input if none provided
if ( !m_settings->HasInputBamFilename )
m_settings->InputFiles.push_back(Options::StandardIn());
// opens the BAM files (by default without checking for indexes)
BamMultiReader reader;
if ( !reader.Open(m_settings->InputFiles, false, true) ) {
cerr << "ERROR: Could not open input BAM file(s)... Aborting." << endl;
return 1;
}
// retrieve header & reference dictionary info
std::string mergedHeader = reader.GetHeaderText();
RefVector references = reader.GetReferenceData();
// open writer
BamWriter writer;
bool writeUncompressed = ( m_settings->OutputFilename == Options::StandardOut() && !m_settings->IsForceCompression );
if ( !writer.Open(m_settings->OutputFilename, mergedHeader, references, writeUncompressed) ) {
cerr << "ERROR: Could not open BAM file " << m_settings->OutputFilename << " for writing... Aborting." << endl;
reader.Close();
return 1;
}
// if no region specified, store entire contents of file(s)
if ( !m_settings->HasRegion ) {
BamAlignment al;
while ( reader.GetNextAlignmentCore(al) )
writer.SaveAlignment(al);
}
// otherwise attempt to use region as constraint
else {
// if region string parses OK
BamRegion region;
if ( Utilities::ParseRegionString(m_settings->Region, reader, region) ) {
// attempt to re-open reader with index files
reader.Close();
bool openedOK = reader.Open(m_settings->InputFiles, true, true );
// if error
if ( !openedOK ) {
cerr << "ERROR: Could not open input BAM file(s)... Aborting." << endl;
return 1;
}
// if index data available, we can use SetRegion
if ( reader.IsIndexLoaded() ) {
// attempt to use SetRegion(), if failed report error
if ( !reader.SetRegion(region.LeftRefID, region.LeftPosition, region.RightRefID, region.RightPosition) ) {
cerr << "ERROR: Region requested, but could not set BamReader region to REGION: " << m_settings->Region << " Aborting." << endl;
reader.Close();
return 1;
}
// everything checks out, just iterate through specified region, storing alignments
BamAlignment al;
while ( reader.GetNextAlignmentCore(al) )
writer.SaveAlignment(al);
}
// no index data available, we have to iterate through until we
// find overlapping alignments
else {
BamAlignment al;
while ( reader.GetNextAlignmentCore(al) ) {
if ( (al.RefID >= region.LeftRefID) && ( (al.Position + al.Length) >= region.LeftPosition ) &&
(al.RefID <= region.RightRefID) && ( al.Position <= region.RightPosition) )
{
writer.SaveAlignment(al);
}
}
}
}
// error parsing REGION string
else {
cerr << "ERROR: Could not parse REGION - " << m_settings->Region << endl;
cerr << "Be sure REGION is in valid format (see README) and that coordinates are valid for selected references" << endl;
reader.Close();
writer.Close();
return 1;
}
}
// clean & exit
reader.Close();
writer.Close();
return 0;
}