/// <summary>
/// The default constructor which chooses the default encoding based on the alphabet.
/// </summary>
protected AbstractBAMParser(string bamFileName, string refSeqName)
: this(bamFileName)
{
// verify whether there is any reads related to chromosome.
var refSeqIndex = RefSeqNames.IndexOf(refSeqName);
if (refSeqIndex < 0)
{
var message = string.Format("reference sequence not found : {0}", refSeqName);
throw new ArgumentException(message, "refSeqName");
}
var bamIndexFileName = GetBAMIndexFileName(bamFileName);
using (var bamIndexFile = new BAMIndexFile(bamIndexFileName, FileMode.Open, FileAccess.Read))
{
var bamIndexInfo = bamIndexFile.Read();
if (bamIndexInfo.RefIndexes.Count < refSeqIndex)
{
throw new ArgumentOutOfRangeException("refSeqIndex");
}
var refIndex = bamIndexInfo.RefIndexes[refSeqIndex];
RefSeqChunks = GetChunks(refIndex);
}
RefSeqName = refSeqName;
ResetChunks();
}
/// <summary>
/// Parses specified BAM file using index file.
/// Index file is assumed to be in the same location as that of the specified bam file with the name "filename".bai
/// For example, if the specified bam file name is D:\BAMdata\sample.bam then index file name will be taken as D:\BAMdata\sample.bam.bai
/// If index file is not available then this method throw an exception.
/// </summary>
/// <param name="fileName">BAM file name.</param>
/// <param name="refSeqIndex">Index of reference sequence.</param>
/// <param name="start">Start index.</param>
/// <param name="end">End index.</param>
/// <returns>SequenceAlignmentMap object which contains alignments overlaps with the specified start
/// and end co-ordinate of the specified reference sequence.</returns>
public SequenceAlignmentMap ParseRange(string fileName, int refSeqIndex, int start, int end)
{
using (FileStream bamStream = new FileStream(fileName, FileMode.Open, FileAccess.Read, FileShare.Read))
{
string bamIndexFileName = getBAMIndexFileName(fileName);
using (BAMIndexFile bamIndexFile = new BAMIndexFile(bamIndexFileName, FileMode.Open, FileAccess.Read))
{
return GetAlignment(bamStream, bamIndexFile, refSeqIndex, start, end);
}
}
}
/// <summary>
/// Parses specified BAM file using index file.
/// </summary>
/// <param name="fileName">BAM file name.</param>
/// <param name="range">SequenceRange object which contains reference sequence name and start and end co-ordinates.</param>
/// <returns>SequenceAlignmentMap object which contains alignments for specified reference sequence and for specified range.</returns>
public SequenceAlignmentMap ParseRange(string fileName, SequenceRange range)
{
if (string.IsNullOrWhiteSpace(fileName))
{
throw new ArgumentNullException("fileName");
}
if (range == null)
{
throw new ArgumentNullException("range");
}
if (string.IsNullOrEmpty(range.ID))
{
throw new ArgumentException("Reference sequence name (range.ID) can't empty or null.");
}
int start = range.Start >= int.MaxValue ? int.MaxValue : (int)range.Start;
int end = range.End >= int.MaxValue ? int.MaxValue : (int)range.End;
if (start == 0 && end == int.MaxValue)
{
using (FileStream bamStream = new FileStream(fileName, FileMode.Open, FileAccess.Read, FileShare.Read))
{
string bamIndexFileName = getBAMIndexFileName(fileName);
using (BAMIndexFile bamIndexFile = new BAMIndexFile(bamIndexFileName, FileMode.Open, FileAccess.Read))
{
return GetAlignment(bamStream, bamIndexFile, range.ID);
}
}
}
else
{
using (FileStream bamStream = new FileStream(fileName, FileMode.Open, FileAccess.Read, FileShare.Read))
{
string bamIndexFileName = getBAMIndexFileName(fileName);
using (BAMIndexFile bamIndexFile = new BAMIndexFile(bamIndexFileName, FileMode.Open, FileAccess.Read))
{
return GetAlignment(bamStream, bamIndexFile, range.ID, start, end);
}
}
}
}
private IEnumerable<SAMAlignedSequence> GetAlignmentYield(Stream bamStream, BAMIndexFile bamIndexFile, int refSeqIndex, int start, int end)
{
foreach (SAMAlignedSequence seq in GetAlignmentMapIterator(bamStream, bamIndexFile, null, refSeqIndex, start, end))
{
yield return seq;
}
}
/// <summary>
/// Parses specified BAM file using index file.
/// Index file is assumed to be in the same location as that of the specified bam file with the name "filename".bai
/// For example, if the specified bam file name is D:\BAMdata\sample.bam then index file name will be taken as D:\BAMdata\sample.bam.bai
/// If index file is not available then this method throw an exception.
/// </summary>
/// <param name="fileName">BAM file name.</param>
/// <param name="refSeqName">Name of reference sequence.</param>
/// <param name="start">Start index.</param>
/// <param name="end">End index.</param>
/// <returns>SequenceAlignmentMap object which contains alignments overlaps with the specified start
/// and end co-ordinate of the specified reference sequence.</returns>
public SequenceAlignmentMap ParseRange(string fileName, string refSeqName, int start, int end)
{
if (refSeqName == null)
{
throw new ArgumentNullException("refSeqName");
}
using (FileStream bamStream = new FileStream(fileName, FileMode.Open, FileAccess.Read, FileShare.Read))
{
string bamIndexFileName = getBAMIndexFileName(fileName);
using (BAMIndexFile bamIndexFile = new BAMIndexFile(bamIndexFileName, FileMode.Open, FileAccess.Read))
{
return GetAlignment(bamStream, bamIndexFile, refSeqName, start, end);
}
}
}
private IEnumerable<SAMAlignedSequence> GetAlignmentWithIndexYield(BAMIndexFile bamIndexFile, int refSeqIndex, int start, int end,
SAMAlignmentHeader header)
{
BAMIndex bamIndexInfo;
BAMReferenceIndexes refIndex;
IList<Chunk> chunks;
bamIndexInfo = bamIndexFile.Read();
if (refSeqIndex != -1 && bamIndexInfo.RefIndexes.Count <= refSeqIndex)
{
throw new ArgumentOutOfRangeException("refSeqIndex");
}
refIndex = bamIndexInfo.RefIndexes[refSeqIndex];
if (start == 0 && end == int.MaxValue)
{
chunks = GetChunks(refIndex);
}
else
{
chunks = GetChunks(refIndex, start, end);
}
IList<SAMAlignedSequence> alignedSeqs = GetAlignedSequences(chunks, start, end);
foreach (SAMAlignedSequence alignedSeq in alignedSeqs)
{
yield return alignedSeqs[0];
}
readStream = null;
}
private IEnumerable<SAMAlignedSequence> GetAlignmentYield(Stream bamStream, BAMIndexFile bamIndexFile, string refSeqName)
{
foreach (SAMAlignedSequence seq in GetAlignmentMapIterator(bamStream, bamIndexFile, refSeqName))
{
yield return seq;
}
}
private void GetAlignmentWithIndex(BAMIndexFile bamIndexFile, int refSeqIndex, int start, int end,
SAMAlignmentHeader header, ref SequenceAlignmentMap seqMap)
{
BAMIndex bamIndexInfo;
BAMReferenceIndexes refIndex;
IList<Chunk> chunks;
seqMap = new SequenceAlignmentMap(header);
bamIndexInfo = bamIndexFile.Read();
if (refSeqIndex != -1 && bamIndexInfo.RefIndexes.Count <= refSeqIndex)
{
throw new ArgumentOutOfRangeException("refSeqIndex");
}
refIndex = bamIndexInfo.RefIndexes[refSeqIndex];
if (start == 0 && end == int.MaxValue)
{
chunks = GetChunks(refIndex);
}
else
{
chunks = GetChunks(refIndex, start, end);
}
IList<SAMAlignedSequence> alignedSeqs = GetAlignedSequences(chunks, start, end);
foreach (SAMAlignedSequence alignedSeq in alignedSeqs)
{
seqMap.QuerySequences.Add(alignedSeq);
}
readStream = null;
}
private IEnumerable<SAMAlignedSequence> GetAlignmentMapIterator(Stream reader, BAMIndexFile bamIndexFile = null,
string refSeqName = null, int? refSeq = null, int start = 0, int end = int.MaxValue)
{
SAMAlignmentHeader header;
if (reader == null || reader.Length == 0)
{
throw new FileFormatException(Properties.Resource.BAM_InvalidBAMFile);
}
readStream = reader;
ValidateReader();
header = GetHeader();
if (refSeq.HasValue && refSeqName == null)
{
// verify whether the chromosome index is there in the header or not.
if (refSeq < 0 || refSeq >= header.ReferenceSequences.Count)
{
throw new ArgumentOutOfRangeException("refSeq");
}
}
else if (refSeqName != null && !refSeq.HasValue)
{
refSeq = refSeqNames.IndexOf(refSeqName);
if (refSeq < 0 || !refSeq.HasValue)
{
string message = string.Format(CultureInfo.InvariantCulture, Properties.Resource.BAM_RefSeqNotFound, refSeqName);
throw new ArgumentException(message, "refSeqName");
}
}
else if (refSeq.HasValue && refSeqName != null)
{
throw new ArgumentException("Received values for params reSeqIndex and refSeqName. Only one parameter can have a value, not both.");
}
if (refSeq.HasValue)
{
if (bamIndexFile != null)
{
foreach (SAMAlignedSequence seq in GetAlignmentWithIndexYield(bamIndexFile, (int)refSeq, start, end, header))
{
yield return seq;
}
}
else
{
throw new ArgumentNullException("refSeqIndex");
}
}
else
{
foreach (SAMAlignedSequence seq in GetAlignmentWithoutIndexYield(header))
{
yield return seq;
}
}
}
// Returns SequenceAlignmentMap by parsing specified BAM stream and BAMIndexFile for the specified reference sequence index.
// this method uses linear index information also.
private SequenceAlignmentMap GetAlignment(Stream bamStream, BAMIndexFile bamIndexFile, int refSeqIndex, int start, int end)
{
return GetAlignmentMap(bamStream, bamIndexFile, null, refSeqIndex, start, end);
}
// Returns SequenceAlignmentMap by parsing specified BAM stream and BAMIndexFile for the specified reference sequence index.
// this method uses linear index information also.
private SequenceAlignmentMap GetAlignment(Stream bamStream, BAMIndexFile bamIndexFile, string refSeqName, int start, int end)
{
return GetAlignmentMap(bamStream, bamIndexFile, refSeqName, -1, start, end);
}
// Returns SequenceAlignmentMap by parsing specified BAM stream and BAMIndexFile for the specified reference sequence name.
private SequenceAlignmentMap GetAlignment(Stream bamStream, BAMIndexFile bamIndexFile, string refSeqName)
{
return GetAlignmentMap(bamStream, bamIndexFile, refSeqName);
}
