void setReadCountInSection( vector<int> & raw_counts, string & chr, int center, SamFile & samIn, SamFileHeader & samHeader, vector<RefSeq*> & REF_SEQ )
{
int st = center - WIN/2;
int ed = center + WIN/2;
bool section_status = samIn.SetReadSection( chr.c_str(), st, ed );
if (!section_status) {
std::cerr << "Warning: Unable to set read section: " << chr << ": " << st << "-" << ed << ". Set section depth = 0!" << std::endl;
return;
}
// proper reads
map<string, vector< DiscPair > > disc_recs; // record where the disc come from
ProperDeck pDeck( REF_SEQ );
SamRecord sam_rec;
while( samIn.ReadRecord(samHeader, sam_rec) ) {
bool pass_qc = PassQC( sam_rec );
if ( !pass_qc )
continue;
if ( sam_rec.getFlag() & 0x2 ) {
if ( sam_rec.getInsertSize() > 0 )
pDeck.Add( sam_rec );
else {
RetrievedIndex rv = pDeck.RetrieveIndex( sam_rec );
int index = getRetrievedIndexToRawCounts( rv );
if (index >= 0) { // for read partially in window, only add clip
if ( sam_rec.get1BasedPosition() < st || sam_rec.get1BasedAlignmentEnd() > ed ) {
if ( index >= 2 )
raw_counts[ index ]++;
}
else
raw_counts[ index ]++;
}
}
}
else { // disc: rec info and wait to reset section later
if ( sam_rec.getReadLength() < 30 )
continue;
string mate_chr = sam_rec.getMateReferenceName();
// check if this one is valid as anchor
DiscPair new_pair( 1, 0, sam_rec, REF_SEQ );
disc_recs[mate_chr].push_back( new_pair );
}
}
// disc reads
for( map<string, vector< DiscPair > >::iterator chr_it = disc_recs.begin(); chr_it != disc_recs.end(); chr_it++ ) {
for( vector< DiscPair >::iterator dp_it = chr_it->second.begin(); dp_it != chr_it->second.end(); dp_it++ ) {
bool section_status = samIn.SetReadSection( chr_it->first.c_str(), dp_it->GetSecondAlignPosition(), dp_it->GetSecondAlignPosition() + WIN );
if (!section_status) {
std::cerr << "ERROR: Unable to set read section: " << chr << ": " << st << "-" << ed << std::endl;
exit(1);
}
SamRecord sam_rec;
while( samIn.ReadRecord(samHeader, sam_rec) ) {
bool pass_qc = PassQC( sam_rec );
if ( !pass_qc )
continue;
if ( !DiscSamPass(sam_rec) )
continue;
int position = sam_rec.get1BasedPosition();
if ( position > dp_it->GetFirstAlignPosition() )
break;
if ( sam_rec.getFlag() & 0x2 )
continue;
bool same_pair = dp_it->IsSamePair( sam_rec );
if ( !same_pair )
continue;
// now add ro raw stats: always use first as anchor
dp_it->AddSecondToPair( sam_rec );
Loci loci = dp_it->GetFirstLoci();
if ( dp_it->FirstValid() ) {
int index = getLociToRawCounts( loci );
raw_counts[ index ]++;
// clear & break
break;
}
}
}
}
}
void testIndex(BamIndex& bamIndex)
{
assert(bamIndex.getNumMappedReads(1) == 2);
assert(bamIndex.getNumUnMappedReads(1) == 0);
assert(bamIndex.getNumMappedReads(0) == 4);
assert(bamIndex.getNumUnMappedReads(0) == 1);
assert(bamIndex.getNumMappedReads(23) == -1);
assert(bamIndex.getNumUnMappedReads(23) == -1);
assert(bamIndex.getNumMappedReads(-1) == 0);
assert(bamIndex.getNumUnMappedReads(-1) == 2);
assert(bamIndex.getNumMappedReads(-2) == -1);
assert(bamIndex.getNumUnMappedReads(-2) == -1);
assert(bamIndex.getNumMappedReads(22) == 0);
assert(bamIndex.getNumUnMappedReads(22) == 0);
// Get the chunks for reference id 1.
Chunk testChunk;
SortedChunkList chunkList;
assert(bamIndex.getChunksForRegion(1, -1, -1, chunkList) == true);
assert(!chunkList.empty());
testChunk = chunkList.pop();
assert(chunkList.empty());
assert(testChunk.chunk_beg == 0x4e7);
assert(testChunk.chunk_end == 0x599);
// Get the chunks for reference id 0.
assert(bamIndex.getChunksForRegion(0, -1, -1, chunkList) == true);
assert(!chunkList.empty());
testChunk = chunkList.pop();
assert(chunkList.empty());
assert(testChunk.chunk_beg == 0x360);
assert(testChunk.chunk_end == 0x4e7);
// Get the chunks for reference id 2.
assert(bamIndex.getChunksForRegion(2, -1, -1, chunkList) == true);
assert(!chunkList.empty());
testChunk = chunkList.pop();
assert(chunkList.empty());
assert(testChunk.chunk_beg == 0x599);
assert(testChunk.chunk_end == 0x5ea);
// Get the chunks for reference id 3.
// There isn't one for this ref id, but still successfully read the file,
// so it should return true, but the list should be empty.
assert(bamIndex.getChunksForRegion(3, -1, -1, chunkList) == true);
assert(chunkList.empty());
// Test reading an indexed bam file.
SamFile inFile;
assert(inFile.OpenForRead("testFiles/sortedBam.bam"));
inFile.setSortedValidation(SamFile::COORDINATE);
assert(inFile.ReadBamIndex("testFiles/sortedBam.bam.bai"));
SamFileHeader samHeader;
assert(inFile.ReadHeader(samHeader));
SamRecord samRecord;
// Test getting num mapped/unmapped reads.
assert(inFile.getNumMappedReadsFromIndex(1) == 2);
assert(inFile.getNumUnMappedReadsFromIndex(1) == 0);
assert(inFile.getNumMappedReadsFromIndex(0) == 4);
assert(inFile.getNumUnMappedReadsFromIndex(0) == 1);
assert(inFile.getNumMappedReadsFromIndex(23) == -1);
assert(inFile.getNumUnMappedReadsFromIndex(23) == -1);
assert(inFile.getNumMappedReadsFromIndex(-1) == 0);
assert(inFile.getNumUnMappedReadsFromIndex(-1) == 2);
assert(inFile.getNumMappedReadsFromIndex(-2) == -1);
assert(inFile.getNumUnMappedReadsFromIndex(-2) == -1);
assert(inFile.getNumMappedReadsFromIndex(22) == 0);
assert(inFile.getNumUnMappedReadsFromIndex(22) == 0);
assert(inFile.getNumMappedReadsFromIndex("2", samHeader) == 2);
assert(inFile.getNumUnMappedReadsFromIndex("2", samHeader) == 0);
assert(inFile.getNumMappedReadsFromIndex("1", samHeader) == 4);
assert(inFile.getNumUnMappedReadsFromIndex("1", samHeader) == 1);
assert(inFile.getNumMappedReadsFromIndex("22", samHeader) == 0);
assert(inFile.getNumUnMappedReadsFromIndex("22", samHeader) == 0);
assert(inFile.getNumMappedReadsFromIndex("", samHeader) == 0);
assert(inFile.getNumUnMappedReadsFromIndex("*", samHeader) == 2);
assert(inFile.getNumMappedReadsFromIndex("unknown", samHeader) == -1);
assert(inFile.getNumUnMappedReadsFromIndex("unknown", samHeader) == -1);
assert(inFile.getNumMappedReadsFromIndex("X", samHeader) == 0);
assert(inFile.getNumUnMappedReadsFromIndex("X", samHeader) == 0);
// Section -1 = Ref *: 2 records (8 & 10 from testSam.sam that is reflected
// in the validation.
assert(inFile.SetReadSection(-1));
assert(inFile.ReadRecord(samHeader, samRecord));
validateRead8(samRecord);
assert(inFile.ReadRecord(samHeader, samRecord));
validateRead10(samRecord);
assert(inFile.ReadRecord(samHeader, samRecord) == false);
// Section 2 = Ref 3: 1 records (9 from testSam.sam that is reflected
// in the validation.
assert(inFile.SetReadSection(2));
assert(inFile.ReadRecord(samHeader, samRecord));
validateRead9(samRecord);
assert(inFile.ReadRecord(samHeader, samRecord) == false);
// Section 0 = Ref 1: 5 records (3, 4, 1, 2, & 6 from testSam.sam that is
// reflected in the validation.
assert(inFile.SetReadSection(0));
assert(inFile.ReadRecord(samHeader, samRecord));
validateRead3(samRecord);
assert(inFile.ReadRecord(samHeader, samRecord));
validateRead4(samRecord);
assert(inFile.ReadRecord(samHeader, samRecord));
validateRead1(samRecord);
assert(inFile.ReadRecord(samHeader, samRecord));
validateRead2(samRecord);
assert(inFile.ReadRecord(samHeader, samRecord));
validateRead6(samRecord);
assert(inFile.ReadRecord(samHeader, samRecord) == false);
// Section 1 = Ref 2: 2 records (5 & 7 from testSam.sam that is reflected
// in the validation.
assert(inFile.SetReadSection(1));
assert(inFile.ReadRecord(samHeader, samRecord));
validateRead5(samRecord);
assert(inFile.ReadRecord(samHeader, samRecord));
validateRead7(samRecord);
assert(inFile.ReadRecord(samHeader, samRecord) == false);
// Section 3 to 22 (ref 4 - 23): 0 records.
for(int i = 3; i < 23; i++)
{
assert(inFile.SetReadSection(i));
assert(inFile.ReadRecord(samHeader, samRecord) == false);
}
// Set the read section.
assert(inFile.SetReadSection("1", 1010, 1012));
assert(inFile.ReadRecord(samHeader, samRecord));
validateRead1(samRecord);
assert(inFile.GetNumOverlaps(samRecord) == 2);
assert(samRecord.getNumOverlaps(1010, 1012) == 2);
assert(samRecord.getNumOverlaps(1010, 1020) == 5);
assert(samRecord.getNumOverlaps(1010, 1011) == 1);
assert(samRecord.getNumOverlaps(1011, 1012) == 1);
assert(inFile.ReadRecord(samHeader, samRecord));
validateRead2(samRecord);
assert(inFile.GetNumOverlaps(samRecord) == 0);
assert(samRecord.getNumOverlaps(1010, 1012) == 0);
assert(samRecord.getNumOverlaps(1010, 1020) == 0);
assert(samRecord.getNumOverlaps(1010, 1011) == 0);
assert(samRecord.getNumOverlaps(1011, 1012) == 0);
assert(inFile.ReadRecord(samHeader, samRecord) == false);
assert(inFile.SetReadSection("1", 1010, 1020));
assert(inFile.ReadRecord(samHeader, samRecord));
validateRead1(samRecord);
assert(inFile.GetNumOverlaps(samRecord) == 5);
assert(samRecord.getNumOverlaps(1010, 1012) == 2);
assert(samRecord.getNumOverlaps(1010, 1020) == 5);
assert(samRecord.getNumOverlaps(1010, 1011) == 1);
assert(samRecord.getNumOverlaps(1011, 1012) == 1);
assert(inFile.ReadRecord(samHeader, samRecord));
validateRead2(samRecord);
assert(inFile.GetNumOverlaps(samRecord) == 0);
assert(samRecord.getNumOverlaps(1010, 1012) == 0);
assert(samRecord.getNumOverlaps(1010, 1020) == 0);
assert(samRecord.getNumOverlaps(1010, 1011) == 0);
assert(samRecord.getNumOverlaps(1011, 1012) == 0);
assert(inFile.ReadRecord(samHeader, samRecord) == false);
assert(inFile.SetReadSection("1", 1010, 1011));
assert(inFile.ReadRecord(samHeader, samRecord));
validateRead1(samRecord);
assert(inFile.GetNumOverlaps(samRecord) == 1);
assert(samRecord.getNumOverlaps(1010, 1012) == 2);
assert(samRecord.getNumOverlaps(1010, 1020) == 5);
assert(samRecord.getNumOverlaps(1010, 1011) == 1);
assert(samRecord.getNumOverlaps(1011, 1012) == 1);
assert(inFile.ReadRecord(samHeader, samRecord) == false);
assert(inFile.SetReadSection("1", 1011, 1012));
assert(inFile.ReadRecord(samHeader, samRecord));
validateRead1(samRecord);
assert(inFile.GetNumOverlaps(samRecord) == 1);
assert(samRecord.getNumOverlaps(1010, 1012) == 2);
assert(samRecord.getNumOverlaps(1010, 1020) == 5);
assert(samRecord.getNumOverlaps(1010, 1011) == 1);
assert(samRecord.getNumOverlaps(1011, 1012) == 1);
assert(inFile.ReadRecord(samHeader, samRecord));
validateRead2(samRecord);
assert(inFile.GetNumOverlaps(samRecord) == 0);
assert(samRecord.getNumOverlaps(1010, 1012) == 0);
assert(samRecord.getNumOverlaps(1010, 1020) == 0);
assert(samRecord.getNumOverlaps(1010, 1011) == 0);
assert(samRecord.getNumOverlaps(1011, 1012) == 0);
assert(inFile.ReadRecord(samHeader, samRecord) == false);
}
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;
}
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);
}
bool Stats::getNextSection(SamFile &samIn)
{
static bool alreadyRead = false;
if(myRegionList == NULL)
{
// no region list is set, so just read once.
if(alreadyRead)
{
// No regions and it has already been read, so
// return false, no more to read.
return(false);
}
// Return true that there is more to read, but
// set the flag that it has already been read
// so the next call will return false.
alreadyRead = true;
return(true);
}
else
{
// There is a region list, so read process that.
// Track whether or not a section has been found.
bool sectionFound = false;
myStartPos = 0;
myEndPos = 0;
// Loop until the end of the file or the end of the file or
// a section is found.
while(!sectionFound && !ifeof(myRegionList))
{
myRegBuffer.Clear();
myRegBuffer.ReadLine(myRegionList);
if(myRegBuffer.IsEmpty())
{
// Nothing read, so continue to the next line.
continue;
}
// A line was read, so parse it.
myRegColumn.ReplaceColumns(myRegBuffer, '\t');
if(myRegColumn.Length() < 3)
{
// Incorrectly formatted line.
std::cerr << "Improperly formatted reg line: "
<< myRegBuffer
<< "; Skipping to the next line.\n";
continue;
}
// Check the columns.
if(!myRegColumn[1].AsInteger(myStartPos))
{
// The start position (2nd column) is not an integer.
std::cerr << "Improperly formatted region line, start position "
<< "(2nd column) is not an integer: "
<< myRegColumn[1]
<< "; Skipping to the next line.\n";
}
else if(!myRegColumn[2].AsInteger(myEndPos))
{
// The end position (3rd column) is not an integer.
std::cerr << "Improperly formatted region line, end position "
<< "(3rd column) is not an integer: "
<< myRegColumn[2]
<< "; Skipping to the next line.\n";
}
else if((myStartPos >= myEndPos) && (myEndPos != -1))
{
// The start position is >= the end position
std::cerr << "Improperly formatted region line, the start position "
<< "is >= end position: "
<< myRegColumn[1]
<< " >= "
<< myRegColumn[2]
<< "; Skipping to the next line.\n";
}
else
{
sectionFound = true;
samIn.SetReadSection(myRegColumn[0].c_str(), myStartPos, myEndPos);
}
}
return(sectionFound);
}
}
/*
if a discordant read is mapped to MEI (overlap with MEI coord)
add centor of ( anchor_end + 3*avr_ins_var )
skip unmap & clip
check 3 types at the same time
*/
void Sites::AddDiscoverSiteFromSingleBam( SingleSampleInfo & si )
{
/*for(int i=0;i<NMEI; i++) {
std::cout << "m=" << i << ": ";
for(map<string, map<int, bool> >::iterator t=meiCoord[m].begin(); t!=meiCoord[m].end(); t++)
std::cout << t->first << "->" << t->second.size() << " ";
std::cout << std::endl;
}*/
avr_read_length = si.avr_read_length;
avr_ins_size = si.avr_ins_size;
min_read_length = avr_read_length / 3;
current_depth = si.depth;
// total_depth += current_depth;
resetDepthAddFlag();
SamFile bam;
SamFileHeader bam_header;
OpenBamAndBai( bam,bam_header, si.bam_name );
for( int i=0; i<pchr_list->size(); i++ ) {
string chr = (*pchr_list)[i];
// if ( !single_chr.empty() && chr.compare(single_chr)!=0 )
// continue;
if ( siteList.find(chr) == siteList.end() )
siteList[chr].clear();
// map<string, map<int, SingleSite> >::iterator pchr = siteList[m].find(chr);
// map<string, map<int, bool> >::iterator coord_chr_ptr = meiCoord[m].find(chr);
// if (coord_chr_ptr == meiCoord[m].end())
// continue;
bool section_status;
if (range_start<0) { // no range
section_status = bam.SetReadSection( chr.c_str() );
if (!section_status) {
string str = "Cannot set read section at chr " + chr;
morphWarning( str );
}
}
else { // set range
section_status = bam.SetReadSection( chr.c_str(), range_start, range_end );
if (!section_status) {
string str = "Cannot set read section at chr " + chr + " " + std::to_string(range_start) + "-" + std::to_string(range_end);
morphWarning( str );
}
}
// DO ADDING
// if (siteList[chr].empty())
// p_reach_last = 1;
// else {
// p_reach_last = 0;
pnearest = siteList[chr].begin();
// }
SingleSite new_site; // temporary cluster. will be added to map later.
new_site.depth = current_depth;
bool start_new = 1; // check if need to add new_site to map and start new new_site
SamRecord rec;
int between = 0; // count #reads after new_site.end. If end changed, add it to rcount and reset to zero
while( bam.ReadRecord( bam_header, rec ) ) {
if (!start_new) {
if (rec.get1BasedPosition() >= new_site.end)
between++;
else
new_site.rcount++;
}
if (rec.getFlag() & 0x2)
continue;
if ( OneEndUnmap( rec.getFlag() ) )
continue;
if ( IsSupplementary(rec.getFlag()) )
continue;
if ( rec.getReadLength() < min_read_length )
continue;
if ( rec.getMapQuality() < MIN_QUALITY )
continue;
if (chr.compare(rec.getMateReferenceName())==0 && rec.getInsertSize() < abs(avr_ins_size*2))
continue;
bool is_mei = 0;
vector<bool> is_in_coord;
is_in_coord.resize(3, 0);
for(int m=0; m<NMEI; m++) {
map<string, map<int, bool> >::iterator coord_chr_ptr = meiCoord[m].find(rec.getMateReferenceName());
if (coord_chr_ptr == meiCoord[m].end())
is_in_coord[m] = 0;
else
is_in_coord[m] = isWithinCoord( rec.get1BasedMatePosition(), coord_chr_ptr->second ); // within MEI coord
if (is_in_coord[m])
is_mei = 1;
}
if (!is_mei)
continue;
if (start_new) {
setNewCluster( is_in_coord, new_site,rec);
start_new = 0;
between = 0;
}
else { // add to existing cluster
if ( rec.get1BasedPosition() > new_site.end + avr_ins_size ) { // start new coord
addClusterToMap(new_site, siteList[chr]);
setNewCluster( is_in_coord, new_site, rec);
start_new = 0;
between = 0;
}
else {
addToCurrentCluster( is_in_coord, new_site, rec);
new_site.rcount += between;
between = 0;
}
}
}
// add last one
if (!start_new)
addClusterToMap(new_site, siteList[chr]);
}
bam.Close();
}