--- /dev/null
+// ***************************************************************************\r
+// BamWriter.cpp (c) 2009 Michael Str�mberg, Derek Barnett\r
+// Marth Lab, Department of Biology, Boston College\r
+// All rights reserved.\r
+// ---------------------------------------------------------------------------\r
+// Last modified: 17 August 2010 (DB)\r
+// ---------------------------------------------------------------------------\r
+// Uses BGZF routines were adapted from the bgzf.c code developed at the Broad\r
+// Institute.\r
+// ---------------------------------------------------------------------------\r
+// Provides the basic functionality for producing BAM files\r
+// ***************************************************************************\r
+\r
+#include <iostream>\r
+\r
+#include "BGZF.h"\r
+#include "BamWriter.h"\r
+using namespace BamTools;\r
+using namespace std;\r
+\r
+struct BamWriter::BamWriterPrivate {\r
+\r
+ // data members\r
+ BgzfData mBGZF;\r
+ bool IsBigEndian;\r
+ \r
+ // constructor / destructor\r
+ BamWriterPrivate(void) { \r
+ IsBigEndian = SystemIsBigEndian(); \r
+ }\r
+ \r
+ ~BamWriterPrivate(void) {\r
+ mBGZF.Close();\r
+ }\r
+\r
+ // "public" interface\r
+ void Close(void);\r
+ bool Open(const string& filename, const string& samHeader, const RefVector& referenceSequences, bool isWriteUncompressed);\r
+ void SaveAlignment(const BamAlignment& al);\r
+\r
+ // internal methods\r
+ const unsigned int CalculateMinimumBin(const int begin, int end) const;\r
+ void CreatePackedCigar(const vector<CigarOp>& cigarOperations, string& packedCigar);\r
+ void EncodeQuerySequence(const string& query, string& encodedQuery);\r
+};\r
+\r
+// -----------------------------------------------------\r
+// BamWriter implementation\r
+// -----------------------------------------------------\r
+\r
+// constructor\r
+BamWriter::BamWriter(void) {\r
+ d = new BamWriterPrivate;\r
+}\r
+\r
+// destructor\r
+BamWriter::~BamWriter(void) {\r
+ delete d;\r
+ d = 0;\r
+}\r
+\r
+// closes the alignment archive\r
+void BamWriter::Close(void) { \r
+ d->Close(); \r
+}\r
+\r
+// opens the alignment archive\r
+bool BamWriter::Open(const string& filename, const string& samHeader, const RefVector& referenceSequences, bool isWriteUncompressed) {\r
+ return d->Open(filename, samHeader, referenceSequences, isWriteUncompressed);\r
+}\r
+\r
+// saves the alignment to the alignment archive\r
+void BamWriter::SaveAlignment(const BamAlignment& al) { \r
+ d->SaveAlignment(al);\r
+}\r
+\r
+// -----------------------------------------------------\r
+// BamWriterPrivate implementation\r
+// -----------------------------------------------------\r
+\r
+// closes the alignment archive\r
+void BamWriter::BamWriterPrivate::Close(void) {\r
+ mBGZF.Close();\r
+}\r
+\r
+// calculates minimum bin for a BAM alignment interval\r
+const unsigned int BamWriter::BamWriterPrivate::CalculateMinimumBin(const int begin, int end) const { \r
+ --end;\r
+ if( (begin >> 14) == (end >> 14) ) return 4681 + (begin >> 14);\r
+ if( (begin >> 17) == (end >> 17) ) return 585 + (begin >> 17);\r
+ if( (begin >> 20) == (end >> 20) ) return 73 + (begin >> 20);\r
+ if( (begin >> 23) == (end >> 23) ) return 9 + (begin >> 23);\r
+ if( (begin >> 26) == (end >> 26) ) return 1 + (begin >> 26);\r
+ return 0;\r
+}\r
+\r
+// creates a cigar string from the supplied alignment\r
+void BamWriter::BamWriterPrivate::CreatePackedCigar(const vector<CigarOp>& cigarOperations, string& packedCigar) {\r
+\r
+ // initialize\r
+ const unsigned int numCigarOperations = cigarOperations.size();\r
+ packedCigar.resize(numCigarOperations * BT_SIZEOF_INT);\r
+\r
+ // pack the cigar data into the string\r
+ unsigned int* pPackedCigar = (unsigned int*)packedCigar.data();\r
+\r
+ unsigned int cigarOp;\r
+ vector<CigarOp>::const_iterator coIter;\r
+ for(coIter = cigarOperations.begin(); coIter != cigarOperations.end(); ++coIter) {\r
+\r
+ switch(coIter->Type) {\r
+ case 'M':\r
+ cigarOp = BAM_CMATCH;\r
+ break;\r
+ case 'I':\r
+ cigarOp = BAM_CINS;\r
+ break;\r
+ case 'D':\r
+ cigarOp = BAM_CDEL;\r
+ break;\r
+ case 'N':\r
+ cigarOp = BAM_CREF_SKIP;\r
+ break;\r
+ case 'S':\r
+ cigarOp = BAM_CSOFT_CLIP;\r
+ break;\r
+ case 'H':\r
+ cigarOp = BAM_CHARD_CLIP;\r
+ break;\r
+ case 'P':\r
+ cigarOp = BAM_CPAD;\r
+ break;\r
+ default:\r
+ printf("ERROR: Unknown cigar operation found: %c\n", coIter->Type);\r
+ exit(1);\r
+ }\r
+\r
+ *pPackedCigar = coIter->Length << BAM_CIGAR_SHIFT | cigarOp;\r
+ pPackedCigar++;\r
+ }\r
+}\r
+\r
+// encodes the supplied query sequence into 4-bit notation\r
+void BamWriter::BamWriterPrivate::EncodeQuerySequence(const string& query, string& encodedQuery) {\r
+\r
+ // prepare the encoded query string\r
+ const unsigned int queryLen = query.size();\r
+ const unsigned int encodedQueryLen = (unsigned int)((queryLen / 2.0) + 0.5);\r
+ encodedQuery.resize(encodedQueryLen);\r
+ char* pEncodedQuery = (char*)encodedQuery.data();\r
+ const char* pQuery = (const char*)query.data();\r
+\r
+ unsigned char nucleotideCode;\r
+ bool useHighWord = true;\r
+\r
+ while(*pQuery) {\r
+\r
+ switch(*pQuery) {\r
+ \r
+ case '=':\r
+ nucleotideCode = 0;\r
+ break;\r
+ \r
+ case 'A':\r
+ nucleotideCode = 1;\r
+ break;\r
+ \r
+ case 'C':\r
+ nucleotideCode = 2;\r
+ break;\r
+ \r
+ case 'G':\r
+ nucleotideCode = 4;\r
+ break;\r
+ \r
+ case 'T':\r
+ nucleotideCode = 8;\r
+ break;\r
+ \r
+ case 'N':\r
+ nucleotideCode = 15;\r
+ break;\r
+ \r
+ default:\r
+ printf("ERROR: Only the following bases are supported in the BAM format: {=, A, C, G, T, N}. Found [%c]\n", *pQuery);\r
+ exit(1);\r
+ }\r
+\r
+ // pack the nucleotide code\r
+ if(useHighWord) {\r
+ *pEncodedQuery = nucleotideCode << 4;\r
+ useHighWord = false;\r
+ } else {\r
+ *pEncodedQuery |= nucleotideCode;\r
+ pEncodedQuery++;\r
+ useHighWord = true;\r
+ }\r
+\r
+ // increment the query position\r
+ pQuery++;\r
+ }\r
+}\r
+\r
+// opens the alignment archive\r
+bool BamWriter::BamWriterPrivate::Open(const string& filename, const string& samHeader, const RefVector& referenceSequences, bool isWriteUncompressed) {\r
+\r
+ // open the BGZF file for writing, return failure if error\r
+ if ( !mBGZF.Open(filename, "wb", isWriteUncompressed) )\r
+ return false;\r
+\r
+ // ================\r
+ // write the header\r
+ // ================\r
+\r
+ // write the BAM signature\r
+ const unsigned char SIGNATURE_LENGTH = 4;\r
+ const char* BAM_SIGNATURE = "BAM\1";\r
+ mBGZF.Write(BAM_SIGNATURE, SIGNATURE_LENGTH);\r
+\r
+ // write the SAM header text length\r
+ uint32_t samHeaderLen = samHeader.size();\r
+ if (IsBigEndian) SwapEndian_32(samHeaderLen);\r
+ mBGZF.Write((char*)&samHeaderLen, BT_SIZEOF_INT);\r
+\r
+ // write the SAM header text\r
+ if(samHeaderLen > 0) \r
+ mBGZF.Write(samHeader.data(), samHeaderLen);\r
+\r
+ // write the number of reference sequences\r
+ uint32_t numReferenceSequences = referenceSequences.size();\r
+ if (IsBigEndian) SwapEndian_32(numReferenceSequences);\r
+ mBGZF.Write((char*)&numReferenceSequences, BT_SIZEOF_INT);\r
+\r
+ // =============================\r
+ // write the sequence dictionary\r
+ // =============================\r
+\r
+ RefVector::const_iterator rsIter;\r
+ for(rsIter = referenceSequences.begin(); rsIter != referenceSequences.end(); rsIter++) {\r
+\r
+ // write the reference sequence name length\r
+ uint32_t referenceSequenceNameLen = rsIter->RefName.size() + 1;\r
+ if (IsBigEndian) SwapEndian_32(referenceSequenceNameLen);\r
+ mBGZF.Write((char*)&referenceSequenceNameLen, BT_SIZEOF_INT);\r
+\r
+ // write the reference sequence name\r
+ mBGZF.Write(rsIter->RefName.c_str(), referenceSequenceNameLen);\r
+\r
+ // write the reference sequence length\r
+ int32_t referenceLength = rsIter->RefLength;\r
+ if (IsBigEndian) SwapEndian_32(referenceLength);\r
+ mBGZF.Write((char*)&referenceLength, BT_SIZEOF_INT);\r
+ }\r
+ \r
+ // return success\r
+ return true;\r
+}\r
+\r
+// saves the alignment to the alignment archive\r
+void BamWriter::BamWriterPrivate::SaveAlignment(const BamAlignment& al) {\r
+\r
+ // if BamAlignment contains only the core data and a raw char data buffer\r
+ // (as a result of BamReader::GetNextAlignmentCore())\r
+ if ( al.SupportData.HasCoreOnly ) {\r
+ \r
+ // write the block size\r
+ unsigned int blockSize = al.SupportData.BlockLength;\r
+ if (IsBigEndian) SwapEndian_32(blockSize);\r
+ mBGZF.Write((char*)&blockSize, BT_SIZEOF_INT);\r
+\r
+ // assign the BAM core data\r
+ uint32_t buffer[8];\r
+ buffer[0] = al.RefID;\r
+ buffer[1] = al.Position;\r
+ buffer[2] = (al.Bin << 16) | (al.MapQuality << 8) | al.SupportData.QueryNameLength;\r
+ buffer[3] = (al.AlignmentFlag << 16) | al.SupportData.NumCigarOperations;\r
+ buffer[4] = al.SupportData.QuerySequenceLength;\r
+ buffer[5] = al.MateRefID;\r
+ buffer[6] = al.MatePosition;\r
+ buffer[7] = al.InsertSize;\r
+ \r
+ // swap BAM core endian-ness, if necessary\r
+ if ( IsBigEndian ) { \r
+ for ( int i = 0; i < 8; ++i )\r
+ SwapEndian_32(buffer[i]); \r
+ }\r
+ \r
+ // write the BAM core\r
+ mBGZF.Write((char*)&buffer, BAM_CORE_SIZE);\r
+ \r
+ // write the raw char data\r
+ mBGZF.Write((char*)al.SupportData.AllCharData.data(), al.SupportData.BlockLength-BAM_CORE_SIZE); \r
+ }\r
+ \r
+ // otherwise, BamAlignment should contain character in the standard fields: Name, QueryBases, etc\r
+ // ( resulting from BamReader::GetNextAlignment() *OR* being generated directly by client code )\r
+ else {\r
+ \r
+ // calculate char lengths\r
+ const unsigned int nameLength = al.Name.size() + 1;\r
+ const unsigned int numCigarOperations = al.CigarData.size();\r
+ const unsigned int queryLength = al.QueryBases.size();\r
+ const unsigned int tagDataLength = al.TagData.size();\r
+ \r
+ // no way to tell if BamAlignment.Bin is already defined (no default, invalid value)\r
+ // force calculation of Bin before storing\r
+ const int endPosition = al.GetEndPosition();\r
+ const unsigned int alignmentBin = CalculateMinimumBin(al.Position, endPosition);\r
+ \r
+ // create our packed cigar string\r
+ string packedCigar;\r
+ CreatePackedCigar(al.CigarData, packedCigar);\r
+ const unsigned int packedCigarLength = packedCigar.size();\r
+\r
+ // encode the query\r
+ string encodedQuery;\r
+ EncodeQuerySequence(al.QueryBases, encodedQuery);\r
+ const unsigned int encodedQueryLength = encodedQuery.size(); \r
+ \r
+ // write the block size\r
+ const unsigned int dataBlockSize = nameLength + packedCigarLength + encodedQueryLength + queryLength + tagDataLength;\r
+ unsigned int blockSize = BAM_CORE_SIZE + dataBlockSize;\r
+ if (IsBigEndian) SwapEndian_32(blockSize);\r
+ mBGZF.Write((char*)&blockSize, BT_SIZEOF_INT);\r
+\r
+ // assign the BAM core data\r
+ uint32_t buffer[8];\r
+ buffer[0] = al.RefID;\r
+ buffer[1] = al.Position;\r
+ buffer[2] = (alignmentBin << 16) | (al.MapQuality << 8) | nameLength;\r
+ buffer[3] = (al.AlignmentFlag << 16) | numCigarOperations;\r
+ buffer[4] = queryLength;\r
+ buffer[5] = al.MateRefID;\r
+ buffer[6] = al.MatePosition;\r
+ buffer[7] = al.InsertSize;\r
+ \r
+ // swap BAM core endian-ness, if necessary\r
+ if ( IsBigEndian ) { \r
+ for ( int i = 0; i < 8; ++i )\r
+ SwapEndian_32(buffer[i]); \r
+ }\r
+ \r
+ // write the BAM core\r
+ mBGZF.Write((char*)&buffer, BAM_CORE_SIZE);\r
+ \r
+ // write the query name\r
+ mBGZF.Write(al.Name.c_str(), nameLength);\r
+\r
+ // write the packed cigar\r
+ if ( IsBigEndian ) {\r
+ \r
+ char* cigarData = (char*)calloc(sizeof(char), packedCigarLength);\r
+ memcpy(cigarData, packedCigar.data(), packedCigarLength);\r
+ \r
+ for (unsigned int i = 0; i < packedCigarLength; ++i) {\r
+ if ( IsBigEndian )\r
+ SwapEndian_32p(&cigarData[i]); \r
+ }\r
+ \r
+ mBGZF.Write(cigarData, packedCigarLength);\r
+ free(cigarData); \r
+ } \r
+ else \r
+ mBGZF.Write(packedCigar.data(), packedCigarLength);\r
+\r
+ // write the encoded query sequence\r
+ mBGZF.Write(encodedQuery.data(), encodedQueryLength);\r
+\r
+ // write the base qualities\r
+ string baseQualities(al.Qualities);\r
+ char* pBaseQualities = (char*)al.Qualities.data();\r
+ for(unsigned int i = 0; i < queryLength; i++) { \r
+ pBaseQualities[i] -= 33; \r
+ }\r
+ mBGZF.Write(pBaseQualities, queryLength);\r
+\r
+ // write the read group tag\r
+ if ( IsBigEndian ) {\r
+ \r
+ char* tagData = (char*)calloc(sizeof(char), tagDataLength);\r
+ memcpy(tagData, al.TagData.data(), tagDataLength);\r
+ \r
+ int i = 0;\r
+ while ( (unsigned int)i < tagDataLength ) {\r
+ \r
+ i += 2; // skip tag type (e.g. "RG", "NM", etc)\r
+ uint8_t type = toupper(tagData[i]); // lower & upper case letters have same meaning \r
+ ++i; // skip value type\r
+ \r
+ switch (type) {\r
+ \r
+ case('A') :\r
+ case('C') : \r
+ ++i;\r
+ break;\r
+ \r
+ case('S') : \r
+ SwapEndian_16p(&tagData[i]); \r
+ i+=2; // sizeof(uint16_t)\r
+ break;\r
+ \r
+ case('F') :\r
+ case('I') : \r
+ SwapEndian_32p(&tagData[i]);\r
+ i+=4; // sizeof(uint32_t)\r
+ break;\r
+ \r
+ case('D') : \r
+ SwapEndian_64p(&tagData[i]);\r
+ i+=8; // sizeof(uint64_t)\r
+ break;\r
+ \r
+ case('H') :\r
+ case('Z') : \r
+ while (tagData[i]) { ++i; }\r
+ ++i; // increment one more for null terminator\r
+ break;\r
+ \r
+ default : \r
+ printf("ERROR: Invalid tag value type\n"); // shouldn't get here\r
+ free(tagData);\r
+ exit(1); \r
+ }\r
+ }\r
+ \r
+ mBGZF.Write(tagData, tagDataLength);\r
+ free(tagData);\r
+ } \r
+ else \r
+ mBGZF.Write(al.TagData.data(), tagDataLength); \r
+ }\r
+}\r
projects / bamtools.git / blobdiff