// BamAlignment.cpp (c) 2009 Derek Barnett
// Marth Lab, Department of Biology, Boston College
// ---------------------------------------------------------------------------
-// Last modified: 8 October 2011 (DB)
+// Last modified: 17 October 2011 (DB)
// ---------------------------------------------------------------------------
// Provides the BamAlignment data structure
// ***************************************************************************
-#include <api/BamAlignment.h>
-#include <api/BamConstants.h>
+#include "api/BamAlignment.h"
+#include "api/BamConstants.h"
using namespace BamTools;
using namespace std;
*/
BamAlignment::~BamAlignment(void) { }
-///*! \fn bool BamAlignment::AddTag(const std::string& tag, const std::string& type, const std::string& value)
-// \brief Adds a field with string data to the BAM tags.
-
-// Does NOT modify an existing tag - use \link BamAlignment::EditTag() \endlink instead.
-
-// \param[in] tag 2-character tag name
-// \param[in] type 1-character tag type (must be "Z" or "H")
-// \param[in] value string data to store
-// \return \c true if the \b new tag was added successfully
-// \sa \samSpecURL for more details on reserved tag names, supported tag types, etc.
-//*/
-
-
-///*! \fn bool AddTag(const std::string& tag, const std::vector<uint8_t>& values);
-// \brief Adds a numeric array field to the BAM tags.
-
-// Does NOT modify an existing tag - use \link BamAlignment::EditTag() \endlink instead.
-
-// \param tag 2-character tag name
-// \param values vector of uint8_t values to store
-
-// \return \c true if the \b new tag was added successfully
-// \sa \samSpecURL for more details on reserved tag names, supported tag types, etc.
-//*/
-
/*! \fn bool BamAlignment::BuildCharData(void)
\brief Populates alignment string fields (read name, bases, qualities, tag data).
memcpy((char*)(TagData.data()), tagData, tagDataLength);
}
- // clear the core-only flag
+ // clear core-only flag & return success
SupportData.HasCoreOnly = false;
-
- // return success
return true;
}
-///*! \fn bool BamAlignment::EditTag(const std::string& tag, const std::string& type, const std::string& value)
-// \brief Edits a BAM tag field containing string data.
-
-// If \a tag does not exist, a new entry is created.
-
-// \param tag 2-character tag name
-// \param type 1-character tag type (must be "Z" or "H")
-// \param value string data to store
-
-// \return \c true if the tag was modified/created successfully
-
-// \sa BamAlignment::RemoveTag()
-// \sa \samSpecURL for more details on reserved tag names, supported tag types, etc.
-//*/
-
-///*! \fn bool EditTag(const std::string& tag, const std::vector<uint8_t>& values);
-// \brief Edits a BAM tag field containing a numeric array.
-
-// If \a tag does not exist, a new entry is created.
-
-// \param tag 2-character tag name
-// \param value vector of uint8_t values to store
-
-// \return \c true if the tag was modified/created successfully
-// \sa \samSpecURL for more details on reserved tag names, supported tag types, etc.
-//*/
-
-/*! \fn bool BamAlignment::FindTag(const std::string& tag, char*& pTagData, const unsigned int& tagDataLength, unsigned int& numBytesParsed)
+/*! \fn bool BamAlignment::FindTag(const std::string& tag, char*& pTagData, const unsigned int& tagDataLength, unsigned int& numBytesParsed) const
\internal
Searches for requested tag in BAM tag data.
- \param tag requested 2-character tag name
- \param pTagData pointer to current position in BamAlignment::TagData
- \param tagDataLength length of BamAlignment::TagData
- \param numBytesParsed number of bytes parsed so far
+ \param[in] tag requested 2-character tag name
+ \param[in,out] pTagData pointer to current position in BamAlignment::TagData
+ \param[in] tagDataLength length of BamAlignment::TagData
+ \param[in,out] numBytesParsed number of bytes parsed so far
\return \c true if found
return false;
}
-/*! \fn bool BamAlignment::GetEditDistance(uint32_t& editDistance) const
- \brief Retrieves value of edit distance tag ("NM").
-
- \deprecated Instead use BamAlignment::GetTag()
- \code
- BamAlignment::GetTag("NM", editDistance);
- \endcode
-
- \param editDistance destination for retrieved value
-
- \return \c true if found
-*/
-
-// TODO : REMOVE THIS METHOD
-bool BamAlignment::GetEditDistance(uint32_t& editDistance) const {
- return GetTag("NM", (uint32_t&)editDistance);
-}
-
-/*! \fn int BamAlignment::GetEndPosition(bool usePadded = false, bool closedInterval = USE_CLOSED_DEFAULT) const
- \brief Calculates alignment end position, based on starting position and CIGAR data.
+/*! \fn int BamAlignment::GetEndPosition(bool usePadded = false, bool closedInterval = false) const
+ \brief Calculates alignment end position, based on its starting position and CIGAR data.
\warning The position returned now represents a zero-based, HALF-OPEN interval.
In previous versions of BamTools (0.x & 1.x) all intervals were treated
- as zero-based, CLOSED. I whole-heartedly apologize for any inconsistencies this
- may have caused if you assumed that BT was always half-open; full aplogies also
- to those who recognized that BamTools originally used a closed interval, but may
- need to update their code to reflect this new change.
-
- \param usePadded Allow inserted bases to affect the reported position. Default is false, so that
- reported position stays synced with reference coordinates.
+ as zero-based, CLOSED.
- \param closedInterval Setting this to true will return a 0-based end coordinate. Default is false,
- so that his value represents a standard, half-open interval.
+ \param[in] usePadded Allow inserted bases to affect the reported position. Default is
+ false, so that reported position stays synced with reference
+ coordinates.
+ \param[in] closedInterval Setting this to true will return a 0-based end coordinate. Default is
+ false, so that his value represents a standard, half-open interval.
\return alignment end position
*/
alignEnd += op.Length;
break;
- // increase end position when CIGAR char is 'I' only if @usePadded is true
+ // increase end position on insertion, only if @usePadded is true
case Constants::BAM_CIGAR_INS_CHAR :
if ( usePadded )
alignEnd += op.Length;
}
/*! \fn std::string BamAlignment::GetErrorString(void) const
- \brief Returns a description of the last error that occurred
+ \brief Returns a human-readable description of the last error that occurred
- This method allows elimnation of STDERR pollution. Developers of client code
+ This method allows elimination of STDERR pollution. Developers of client code
may choose how the messages are displayed to the user, if at all.
- \return description of last error that occurred
+ \return error description
*/
std::string BamAlignment::GetErrorString(void) const {
return ErrorString;
}
-/*! \fn bool BamAlignment::GetReadGroup(std::string& readGroup) const
- \brief Retrieves value of read group tag ("RG").
+/*! \fn bool BamAlignment::GetSoftClips(std::vector<int>& clipSizes, std::vector<int>& readPositions, std::vector<int>& genomePositions, bool usePadded = false) const
+ \brief Identifies if an alignment has a soft clip. If so, identifies the
+ sizes of the soft clips, as well as their positions in the read and reference.
- \deprecated Instead use BamAlignment::GetTag()
- \code
- BamAlignment::GetTag("RG", readGroup);
- \endcode
+ \param[out] clipSizes vector of the sizes of each soft clip in the alignment
+ \param[out] readPositions vector of the 0-based read locations of each soft clip in the alignment.
+ These positions are basically indexes within the read, not genomic positions.
+ \param[out] genomePositions vector of the 0-based genome locations of each soft clip in the alignment
+ \param[in] usePadded inserted bases affect reported position. Default is false, so that
+ reported position stays 'sync-ed' with reference coordinates.
- \param readGroup destination for retrieved value
-
- \return \c true if found
+ \return \c true if any soft clips were found in the alignment
*/
+bool BamAlignment::GetSoftClips(vector<int>& clipSizes,
+ vector<int>& readPositions,
+ vector<int>& genomePositions,
+ bool usePadded) const
+{
+ // initialize positions & flags
+ int refPosition = Position;
+ int readPosition = 0;
+ bool softClipFound = false;
+ bool firstCigarOp = true;
-// TODO : REMOVE THIS METHOD
-bool BamAlignment::GetReadGroup(std::string& readGroup) const {
- return GetTag("RG", readGroup);
-}
+ // iterate over cigar operations
+ vector<CigarOp>::const_iterator cigarIter = CigarData.begin();
+ vector<CigarOp>::const_iterator cigarEnd = CigarData.end();
+ for ( ; cigarIter != cigarEnd; ++cigarIter) {
+ const CigarOp& op = (*cigarIter);
-///*! \fn bool BamAlignment::GetTag(const std::string& tag, std::string& destination) const
-// \brief Retrieves the string value associated with a BAM tag.
+ switch ( op.Type ) {
-// \param tag 2-character tag name
-// \param destination destination for retrieved value
+ // increase both read & genome positions on CIGAR chars [DMXN=]
+ case Constants::BAM_CIGAR_DEL_CHAR :
+ case Constants::BAM_CIGAR_MATCH_CHAR :
+ case Constants::BAM_CIGAR_MISMATCH_CHAR :
+ case Constants::BAM_CIGAR_REFSKIP_CHAR :
+ case Constants::BAM_CIGAR_SEQMATCH_CHAR :
+ refPosition += op.Length;
+ readPosition += op.Length;
+ break;
-// \return \c true if found
-//*/
+ // increase read position on insertion, genome position only if @usePadded is true
+ case Constants::BAM_CIGAR_INS_CHAR :
+ readPosition += op.Length;
+ if ( usePadded )
+ refPosition += op.Length;
+ break;
-///*! \fn bool BamAlignment::GetTag(const std::string& tag, std::vector<uint32_t>& destination) const
-// \brief Retrieves the numeric array data associated with a BAM tag
+ case Constants::BAM_CIGAR_SOFTCLIP_CHAR :
+
+ softClipFound = true;
+
+ //////////////////////////////////////////////////////////////////////////////
+ // if we are dealing with the *first* CIGAR operation
+ // for this alignment, we increment the read position so that
+ // the read and genome position of the clip are referring to the same base.
+ // For example, in the alignment below, the ref position would be 4, yet
+ // the read position would be 0. Thus, to "sync" the two,
+ // we need to increment the read position by the length of the
+ // soft clip.
+ // Read: ATCGTTTCGTCCCTGC
+ // Ref: GGGATTTCGTCCCTGC
+ // Cigar: SSSSMMMMMMMMMMMM
+ //
+ // NOTE: This only needs to be done if the soft clip is the _first_ CIGAR op.
+ //////////////////////////////////////////////////////////////////////////////
+ if ( firstCigarOp )
+ readPosition += op.Length;
+
+ // track the soft clip's size, read position, and genome position
+ clipSizes.push_back(op.Length);
+ readPositions.push_back(readPosition);
+ genomePositions.push_back(refPosition);
+
+ // any other CIGAR operations have no effect
+ default :
+ break;
+ }
-// \param tag 2-character tag name
-// \param destination destination for retrieved data
+ // clear our "first pass" flag
+ firstCigarOp = false;
+ }
-// \return \c true if found
-//*/
+ // return whether any soft clips found
+ return softClipFound;
+}
/*! \fn bool BamAlignment::GetTagType(const std::string& tag, char& type) const
\brief Retrieves the BAM tag type-code associated with requested tag name.
- \param tag 2-character tag name
- \param type destination for the retrieved (1-character) tag type
+ \param[in] tag 2-character tag name
+ \param[out] type retrieved (1-character) type-code
\return \c true if found
\sa \samSpecURL for more details on reserved tag names, supported tag types, etc.
/*! \fn bool BamAlignment::HasTag(const std::string& tag) const
\brief Returns true if alignment has a record for requested tag.
- \param tag 2-character tag name
+
+ \param[in] tag 2-character tag name
\return \c true if alignment has a record for tag
*/
bool BamAlignment::HasTag(const std::string& tag) const {
return ( (AlignmentFlag & Constants::BAM_ALIGNMENT_READ_2) != 0 );
}
-/*! \fn bool BamAlignment::IsValidSize(const string& tag, const string& type) const
+/*! \fn bool BamAlignment::IsValidSize(const std::string& tag, const std::string& type) const
\internal
Checks that tag name & type strings are expected sizes.
- \a tag should have length
- \a type should have length 1
-
- \param tag BAM tag name
- \param type BAM tag type-code
- \return \c true if both \a tag and \a type are correct sizes
+ \param tag[in] BAM tag name
+ \param type[in] BAM tag type-code
+ \return \c true if both input strings are valid sizes
*/
bool BamAlignment::IsValidSize(const std::string& tag, const std::string& type) const {
return (tag.size() == Constants::BAM_TAG_TAGSIZE) &&
/*! \fn void BamAlignment::RemoveTag(const std::string& tag)
\brief Removes field from BAM tags.
+
+ \param[in] tag 2-character name of field to remove
*/
void BamAlignment::RemoveTag(const std::string& tag) {
\internal
Sets a formatted error string for this alignment.
+
+ \param[in] where class/method where error occurred
+ \param[in] what description of error
*/
void BamAlignment::SetErrorString(const std::string& where, const std::string& what) const {
static const string SEPARATOR = ": ";
else AlignmentFlag |= Constants::BAM_ALIGNMENT_MATE_UNMAPPED;
}
-/*! \fn void BamAlignment::SetIsMateUnmapped(bool ok)
- \brief Complement of using SetIsMateMapped().
- \deprecated For sake of symmetry with the query methods
- \sa IsMateMapped(), SetIsMateMapped()
-*/
-void BamAlignment::SetIsMateUnmapped(bool ok) {
- SetIsMateMapped(!ok);
-}
-
/*! \fn void BamAlignment::SetIsMateReverseStrand(bool ok)
\brief Sets "alignment's mate mapped to reverse strand" flag to \a ok.
*/
else AlignmentFlag &= ~Constants::BAM_ALIGNMENT_REVERSE_STRAND;
}
-/*! \fn void BamAlignment::SetIsSecondaryAlignment(bool ok)
- \brief Complement of using SetIsPrimaryAlignment().
- \deprecated For sake of symmetry with the query methods
- \sa IsPrimaryAlignment(), SetIsPrimaryAlignment()
-*/
-void BamAlignment::SetIsSecondaryAlignment(bool ok) {
- SetIsPrimaryAlignment(!ok);
-}
-
/*! \fn void BamAlignment::SetIsSecondMate(bool ok)
\brief Sets "alignment is second mate on read" flag to \a ok.
*/
else AlignmentFlag &= ~Constants::BAM_ALIGNMENT_READ_2;
}
-/*! \fn void BamAlignment::SetIsUnmapped(bool ok)
- \brief Complement of using SetIsMapped().
- \deprecated For sake of symmetry with the query methods
- \sa IsMapped(), SetIsMapped()
-*/
-void BamAlignment::SetIsUnmapped(bool ok) {
- SetIsMapped(!ok);
-}
-
-/*! \fn bool BamAlignment::SkipToNextTag(const char storageType, char*& pTagData, unsigned int& numBytesParsed)
+/*! \fn bool BamAlignment::SkipToNextTag(const char storageType, char*& pTagData, unsigned int& numBytesParsed) const
\internal
Moves to next available tag in tag data string
- \param storageType BAM tag type-code that determines how far to move cursor
- \param pTagData pointer to current position (cursor) in tag string
- \param numBytesParsed report of how many bytes were parsed (cumulatively)
+ \param[in] storageType BAM tag type-code that determines how far to move cursor
+ \param[in,out] pTagData pointer to current position (cursor) in tag string
+ \param[in,out] numBytesParsed report of how many bytes were parsed (cumulatively)
\return \c if storageType was a recognized BAM tag type
- \post \a pTagData will point to the byte where the next tag data begins.
+
+ \post \a pTagData will point to the byte where the next tag data begins.
\a numBytesParsed will correspond to the cursor's position in the full TagData string.
*/
bool BamAlignment::SkipToNextTag(const char storageType,
projects / bamtools.git / blobdiff