//{{{ void process_intra_chrom_pair(const BamAlignment &curr,
void
SV_Pair::
process_intra_chrom_pair(const BamAlignment &curr,
const RefVector refs,
BamWriter &inter_chrom_reads,
map<string, BamAlignment> &mapped_pairs,
UCSCBins<SV_BreakPoint*> &r_bin,
int weight,
int ev_id,
SV_PairReader *reader)
{
if (curr.RefID == curr.MateRefID) {
process_pair(curr,
refs,
mapped_pairs,
r_bin,
weight,
ev_id,
reader);
} else if (curr.IsMapped() &&
curr.IsMateMapped() &&
(curr.RefID >= 0) &&
(curr.MateRefID >= 0) ) {
BamAlignment al = curr;
string x = reader->get_source_file_name();
al.AddTag("LS","Z",x);
inter_chrom_reads.SaveAlignment(al);
}
}
void TagBam::Tag() {
// open the annotations files for processing;
OpenAnnoFiles();
// open the BAM file
BamReader reader;
BamWriter writer;
if (!reader.Open(_bamFile)) {
cerr << "Failed to open BAM file " << _bamFile << endl;
exit(1);
}
// get header & reference information
string bamHeader = reader.GetHeaderText();
RefVector refs = reader.GetReferenceData();
// set compression mode
BamWriter::CompressionMode compressionMode = BamWriter::Compressed;
// if ( _isUncompressedBam ) compressionMode = BamWriter::Uncompressed;
writer.SetCompressionMode(compressionMode);
// open our BAM writer
writer.Open("stdout", bamHeader, refs);
// rip through the BAM file and test for overlaps with each annotation file.
BamAlignment al;
vector<BED> hits;
while (reader.GetNextAlignment(al)) {
if (al.IsMapped() == true) {
BED a;
a.chrom = refs.at(al.RefID).RefName;
a.start = al.Position;
a.end = al.GetEndPosition(false, false);
a.strand = "+";
if (al.IsReverseStrand()) a.strand = "-";
ostringstream annotations;
// annotate the BAM file based on overlaps with the annotation files.
for (size_t i = 0; i < _annoFiles.size(); ++i)
{
// grab the current annotation file.
BedFile *anno = _annoFiles[i];
if (!_useNames && !_useScores && !_useIntervals) {
// add the label for this annotation file to tag if there is overlap
if (anno->anyHits(a.chrom, a.start, a.end, a.strand,
_sameStrand, _diffStrand, _overlapFraction, false))
{
annotations << _annoLabels[i] << ";";
}
}
// use the score field
else if (!_useNames && _useScores && !_useIntervals) {
anno->allHits(a.chrom, a.start, a.end, a.strand,
hits, _sameStrand, _diffStrand, 0.0, false);
for (size_t i = 0; i < hits.size(); ++i) {
annotations << hits[i].score;
if (i < hits.size() - 1) annotations << ",";
}
if (hits.size() > 0) annotations << ";";
hits.clear();
}
// use the name field from the annotation files to populate tag
else if (_useNames && !_useScores && !_useIntervals) {
anno->allHits(a.chrom, a.start, a.end, a.strand,
hits, _sameStrand, _diffStrand, 0.0, false);
for (size_t j = 0; j < hits.size(); ++j) {
annotations << hits[j].name;
if (j < hits.size() - 1) annotations << ",";
}
if (hits.size() > 0) annotations << ";";
hits.clear();
}
// use the full interval information annotation files to populate tag
else if (!_useNames && !_useScores && _useIntervals) {
anno->allHits(a.chrom, a.start, a.end, a.strand,
hits, _sameStrand, _diffStrand, 0.0, false);
for (size_t j = 0; j < hits.size(); ++j) {
annotations << _annoLabels[i] << ":" <<
hits[j].chrom << ":" <<
hits[j].start << "-" <<
hits[j].end << "," <<
hits[j].name << "," <<
hits[j].score << "," <<
hits[j].strand;
if (j < hits.size() - 1) annotations << ",";
}
if (hits.size() > 0) annotations << ";";
hits.clear();
}
}
// were there any overlaps with which to make a tag?
if (annotations.str().size() > 0) {
al.AddTag(_tag, "Z", annotations.str().substr(0, annotations.str().size() - 1)); // get rid of the last ";"
}
}
writer.SaveAlignment(al);
}
reader.Close();
writer.Close();
// close the annotations files;
CloseAnnoFiles();
}
void ReadContainerTest::test_GetBamTags() {
// TODO
BamAlignment aln;
string str_result;
float float_result;
string rg = "test";
float xc = 12.0;
aln.AddTag("RG","Z",rg);
aln.AddTag("XD","i",0);
aln.AddTag("XC","f",xc);
// GetStringBamTag
CPPUNIT_ASSERT(_read_container->GetStringBamTag(aln, "RG", &str_result));
CPPUNIT_ASSERT(str_result == "test");
CPPUNIT_ASSERT(!_read_container->GetStringBamTag(aln, "XX", &str_result));
// GetFloatBamTag
CPPUNIT_ASSERT(_read_container->GetFloatBamTag(aln, "XC", &float_result));
CPPUNIT_ASSERT(float_result == 12.0);
CPPUNIT_ASSERT(!_read_container->GetFloatBamTag(aln, "XX", &float_result));
// GetIntBamTag
int int_result;
int d = 0;
aln.AddTag("XA","i",d);
CPPUNIT_ASSERT(_read_container->GetIntBamTag(aln, "XA", &int_result));
CPPUNIT_ASSERT_EQUAL(int_result, 0);
aln.RemoveTag("XA");
int8_t d8 = 10;
aln.AddTag("XA","c", d8);
CPPUNIT_ASSERT(_read_container->GetIntBamTag(aln, "XA", &int_result));
CPPUNIT_ASSERT_EQUAL(int_result, 10);
aln.RemoveTag("XA");
aln.AddTag("XA","c",rg);
CPPUNIT_ASSERT(!_read_container->GetFloatBamTag(aln, "XA", &float_result));
aln.RemoveTag("XA");
uint8_t ud8 = 9;
aln.AddTag("XA","C", ud8);
CPPUNIT_ASSERT(_read_container->GetIntBamTag(aln, "XA", &int_result));
CPPUNIT_ASSERT_EQUAL(int_result, 9);
aln.RemoveTag("XA");
aln.AddTag("XA","C",rg);
CPPUNIT_ASSERT(!_read_container->GetFloatBamTag(aln, "XA", &float_result));
aln.RemoveTag("XA");
int16_t d16 = 8;
aln.AddTag("XA","s",d16);
CPPUNIT_ASSERT(_read_container->GetIntBamTag(aln, "XA", &int_result));
CPPUNIT_ASSERT_EQUAL(int_result, 8);
aln.RemoveTag("XA");
aln.AddTag("XA","s",rg);
CPPUNIT_ASSERT(!_read_container->GetFloatBamTag(aln, "XA", &float_result));
aln.RemoveTag("XA");
uint16_t ud16 = 7;
aln.AddTag("XA","S",ud16);
CPPUNIT_ASSERT(_read_container->GetIntBamTag(aln, "XA", &int_result));
CPPUNIT_ASSERT_EQUAL(int_result, 7);
aln.RemoveTag("XA");
aln.AddTag("XA","S",rg);
CPPUNIT_ASSERT(!_read_container->GetFloatBamTag(aln, "XA", &float_result));
aln.RemoveTag("XA");
uint32_t ud32 = 6;
aln.AddTag("XA","I",ud32);
CPPUNIT_ASSERT(_read_container->GetIntBamTag(aln, "XA", &int_result));
CPPUNIT_ASSERT_EQUAL(int_result, 6);
aln.RemoveTag("XA");
aln.AddTag("XA","I",rg);
CPPUNIT_ASSERT(!_read_container->GetFloatBamTag(aln, "XA", &float_result));
aln.RemoveTag("XA");
}
void ReadContainerTest::test_ParseRead() {
map<pair<string,int>, string> ref_ext_nucleotides;
BamAlignment aln;
std::string rg = "test";
std::string repseq = "AC";
AlignedRead aligned_read;
float copynum = 10;
include_flank = false;
// Test valid allele length
aln.Name = "test";
aln.QueryBases = "NNNNN";
aln.Qualities = "NNNNN";
aln.SetIsReverseStrand(true);
aln.Position = 0;
aln.SetIsSecondMate(false);
aln.AddTag("RG","Z",rg);
aln.AddTag("XS","i",0);
aln.AddTag("XE","i",20);
aln.AddTag("XR", "Z", repseq);
aln.AddTag("XC", "f", copynum);
aln.RefID = 0;
// No XD
aln.RemoveTag("XD");
CPPUNIT_ASSERT(!(_read_container->ParseRead(aln, &aligned_read, ref_ext_nucleotides)));
// Test more valid allele lengths
aln.AddTag("XD","i",20);
CPPUNIT_ASSERT(_read_container->ParseRead(aln, &aligned_read, ref_ext_nucleotides));
aln.RemoveTag("XD");
aln.AddTag("XD","i",-19);
CPPUNIT_ASSERT(_read_container->ParseRead(aln, &aligned_read, ref_ext_nucleotides));
// Test invalid allele length
aln.RemoveTag("XD");
aln.AddTag("XD","i",-31);
CPPUNIT_ASSERT(!(_read_container->ParseRead(aln, &aligned_read, ref_ext_nucleotides)));
// Exceed max_diff_ref
aln.RemoveTag("XD");
aln.AddTag("XD","i",100);
CPPUNIT_ASSERT(!(_read_container->ParseRead(aln, &aligned_read, ref_ext_nucleotides)));
aln.RemoveTag("XD");
aln.AddTag("XD","i",0);
// Exceed max mate dist
aln.AddTag("XM","i",1000000);
CPPUNIT_ASSERT(!(_read_container->ParseRead(aln, &aligned_read, ref_ext_nucleotides)));
aln.RemoveTag("XM");
// Exceed max mapq
aln.AddTag("XQ","i",10000);
CPPUNIT_ASSERT(!(_read_container->ParseRead(aln, &aligned_read, ref_ext_nucleotides)));
aln.RemoveTag("XQ");
// Read is mate
aln.SetIsSecondMate(true);
CPPUNIT_ASSERT(!(_read_container->ParseRead(aln, &aligned_read, ref_ext_nucleotides)));
aln.SetIsSecondMate(false);
// Read is partial
aln.AddTag("XP","i",1);
CPPUNIT_ASSERT(!(_read_container->ParseRead(aln, &aligned_read, ref_ext_nucleotides)));
aln.RemoveTag("XP");
// Non-unit
unit = true;
aln.RemoveTag("XD");
aln.AddTag("XD","i",5);
CPPUNIT_ASSERT(!(_read_container->ParseRead(aln, &aligned_read, ref_ext_nucleotides)));
unit = false;
}
void setMateInfo( BamAlignment & rec1, BamAlignment & rec2, SamHeader & header) {
const int NO_ALIGNMENT_REFERENCE_INDEX = -1;
const int NO_ALIGNMENT_START = -1;
// If neither read is unmapped just set their mate info
if (rec1.IsMapped() && rec2.IsMapped()) {
rec1.MateRefID = rec2.MateRefID;
rec1.MatePosition = rec2.Position;
rec1.SetIsReverseStrand(rec2.IsReverseStrand());
rec1.SetIsMapped(true);
rec1.AddTag("MQ", "i", rec2.MapQuality);
rec2.MateRefID = rec1.RefID;
rec2.MatePosition = rec1.Position;
rec2.SetIsReverseStrand( rec1.IsReverseStrand() );
rec2.SetIsMapped(true);
rec2.AddTag("MQ", "i", rec1.MapQuality);
}
// Else if they're both unmapped set that straight
else if (!rec1.IsMapped() && !rec2.IsMapped()) {
rec1.RefID = NO_ALIGNMENT_REFERENCE_INDEX;
rec1.Position = NO_ALIGNMENT_START;
rec1.MateRefID = NO_ALIGNMENT_REFERENCE_INDEX;
rec1.MatePosition = NO_ALIGNMENT_START;
rec1.SetIsReverseStrand(rec2.IsReverseStrand());
rec1.SetIsMapped(false);
rec2.RemoveTag("MQ");
rec1.Length = 0;
rec2.RefID = NO_ALIGNMENT_REFERENCE_INDEX;
rec2.Position = NO_ALIGNMENT_START;
rec2.MateRefID = NO_ALIGNMENT_REFERENCE_INDEX;
rec2.MatePosition = NO_ALIGNMENT_START;
rec2.SetIsReverseStrand(rec1.IsReverseStrand());
rec2.SetIsMapped(false);
rec2.RemoveTag("MQ");
rec2.Length = 0;
}
// And if only one is mapped copy it's coordinate information to the mate
else {
BamAlignment & mapped = rec1.IsMapped() ? rec1 : rec2;
BamAlignment & unmapped = rec1.IsMapped() ? rec2 : rec1;
unmapped.RefID = mapped.RefID;
unmapped.Position = mapped.Position;
mapped.MateRefID = unmapped.RefID;
mapped.MatePosition = unmapped.Position;
mapped.SetIsMateReverseStrand(unmapped.IsReverseStrand());
mapped.SetIsMateMapped(false);
mapped.Length = 0;
unmapped.MateRefID = mapped.RefID;
unmapped.MatePosition = mapped.Position;
unmapped.SetIsMateReverseStrand(mapped.IsReverseStrand());
unmapped.SetIsMateMapped(true);
unmapped.Length = 0;
}
const int insertSize = computeInsertSize(rec1, rec2);
rec1.Length = insertSize;
rec2.Length = -insertSize;
}
