/// <summary>
/// Validate BAM file Header fields.
/// </summary>
/// <param name="nodeName">XML nodename used for different test cases</param>
/// <param name="seqAlignment">seqAlignment object</param>
void ValidateBAMHeaderRecords(string nodeName,
SequenceAlignmentMap seqAlignment)
{
string expectedHeaderTagValues = _utilityObj._xmlUtil.GetTextValue(
nodeName, Constants.RecordTagValuesNode);
string expectedHeaderTagKeys = _utilityObj._xmlUtil.GetTextValue(
nodeName, Constants.RecordTagKeysNode);
string expectedHeaderTypes = _utilityObj._xmlUtil.GetTextValue(
nodeName, Constants.HeaderTyepsNodes);
string[] expectedHeaderTagsValues = expectedHeaderTagValues.Split(',');
string[] expectedHeaderKeys = expectedHeaderTagKeys.Split(',');
string[] expectedHeaders = expectedHeaderTypes.Split(',');
SAMAlignmentHeader header = seqAlignment.Header;
IList <SAMRecordField> recordFields = header.RecordFields;
int tagKeysCount = 0;
int tagValuesCount = 0;
for (int index = 0; index < recordFields.Count; index++)
{
Assert.AreEqual(expectedHeaders[index].Replace("/", ""),
recordFields[index].Typecode.ToString((IFormatProvider)null).Replace("/", ""));
for (int tags = 0; tags < recordFields[index].Tags.Count; tags++)
{
Assert.AreEqual(expectedHeaderKeys[tagKeysCount].Replace("/", ""),
recordFields[index].Tags[tags].Tag.ToString((IFormatProvider)null).Replace("/", ""));
Assert.AreEqual(expectedHeaderTagsValues[tagValuesCount].Replace("/", ""),
recordFields[index].Tags[tags].Value.ToString((IFormatProvider)null).Replace("/", "").Replace("\r", "").Replace("\n", ""));
tagKeysCount++;
tagValuesCount++;
}
}
}
/// <summary>
/// Updates the header with reference name and length from ReferenceNamesAndLength file.
/// </summary>
/// <param name="header">SAM alignment header.</param>
private void UpdateReferenceInformationFromFile(SAMAlignmentHeader header)
{
header.ReferenceSequences.Clear();
using (StreamReader reader = new StreamReader(ReferenceNamesAndLength))
{
header.ReferenceSequences.Clear();
string read = reader.ReadLine();
while (!string.IsNullOrEmpty(read))
{
string[] splitRegion = read.Split(new string[] { "\t" }, StringSplitOptions.RemoveEmptyEntries);
if (splitRegion.Length > 1)
{
string name = splitRegion[0];
long len = long.Parse(splitRegion[1], CultureInfo.InvariantCulture);
header.ReferenceSequences.Add(new ReferenceSequenceInfo(name, len));
}
else
{
throw new InvalidOperationException("Invalid file for reference name and length");
}
read = reader.ReadLine();
}
}
}
/// <summary>
/// Dispalys the headers present in the BAM file
/// </summary>
/// <param name="seqAlignmentMap">SeqAlignment map</param>
private void DisplayHeader(SequenceAlignmentMap seqAlignmentMap)
{
// Get Header
SAMAlignmentHeader header = seqAlignmentMap.Header;
IList <SAMRecordField> recordField = header.RecordFields;
IList <string> commenstList = header.Comments;
if (recordField.Count > 0)
{
Console.WriteLine("MetaData:");
// Read Header Lines
for (int i = 0; i < recordField.Count; i++)
{
Console.Write("\n@{0}", recordField[i].Typecode);
for (int tags = 0; tags < recordField[i].Tags.Count; tags++)
{
Console.Write("\t{0}:{1}", recordField[i].Tags[tags].Tag,
recordField[i].Tags[tags].Value);
}
}
}
// Displays the comments if any
if (commenstList.Count > 0)
{
for (int i = 0; i < commenstList.Count; i++)
{
Console.Write("\n@CO\t{0}\n", commenstList[i].ToString());
}
}
}
public void ValidateSAMParserHeader()
{
string filePath = utilityObj.xmlUtil.GetTextValue(
Constants.SmallSAMFileNode, Constants.FilePathNode);
string[] expectedHeaderTagValues = utilityObj.xmlUtil.GetTextValue(
Constants.SmallSAMFileNode, Constants.RecordTagValuesNode).Split(',');
string[] expectedHeaderTagKeys = utilityObj.xmlUtil.GetTextValue(
Constants.SmallSAMFileNode, Constants.RecordTagKeysNode).Split(',');
string[] expectedHeaderTypes = utilityObj.xmlUtil.GetTextValue(
Constants.SmallSAMFileNode, Constants.HeaderTyepsNodes).Split(',');
SAMAlignmentHeader aligntHeader =
SAMParser.ParseSAMHeader(filePath);
int tagKeysCount = 0;
int tagValuesCount = 0;
for (int index = 0; index < aligntHeader.RecordFields.Count; index++)
{
Assert.AreEqual(expectedHeaderTypes[index].Replace("/", ""),
aligntHeader.RecordFields[index].Typecode.ToString((IFormatProvider)null).Replace("/", ""));
for (int tags = 0; tags < aligntHeader.RecordFields[index].Tags.Count; tags++)
{
Assert.AreEqual(
expectedHeaderTagKeys[tagKeysCount].Replace("/", ""),
aligntHeader.RecordFields[index].Tags[tags].Tag.ToString((IFormatProvider)null).Replace("/", ""));
Assert.AreEqual(
expectedHeaderTagValues[tagValuesCount].Replace("/", ""),
aligntHeader.RecordFields[index].Tags[tags].Value.ToString((IFormatProvider)null).Replace("/", "").Replace("\r", "").Replace("\n", ""));
tagKeysCount++;
tagValuesCount++;
}
}
}
/// <summary>
/// Writes specified sequence alignment to stream.
/// The output is formatted according to the BAM structure.
/// </summary>
/// <param name="sequenceAlignmentMap">SequenceAlignmentMap object.</param>
/// <param name="writer">Stream to write.</param>
/// <param name="createSortedFile">If this flag is true output file will be sorted.</param>
private void WriteUncompressed(SequenceAlignmentMap sequenceAlignmentMap, Stream writer, bool createSortedFile)
{
SAMAlignmentHeader header = sequenceAlignmentMap.Header;
if (createSortedFile && SortType == BAMSortByFields.ChromosomeNameAndCoordinates)
{
header = GetHeaderWithSortedSQFields(header, true);
this.refSequences = header.GetReferenceSequenceRanges();
}
if (this.refSequences == null)
{
this.refSequences = header.GetReferenceSequenceRanges();
}
WriteHeader(header, writer);
writer.Flush();
if (createSortedFile)
{
WriteUncompressedSortedBAM(sequenceAlignmentMap, writer);
}
else
{
foreach (SAMAlignedSequence seq in sequenceAlignmentMap.QuerySequences)
{
SAMAlignedSequence alignedSeq = seq;
this.ValidateSQHeader(alignedSeq.RName);
this.WriteAlignedSequence(alignedSeq, writer);
writer.Flush();
}
}
writer.Flush();
}
/// <summary>
/// Converts the input BAM to SAM file format.
/// </summary>
private void ConvertFromBAMToSAM()
{
using (Stream stream = new FileStream(InputFilePath, FileMode.Open, FileAccess.Read))
{
SAMAlignmentHeader header = null;
try
{
header = bamparser.GetHeader(stream);
}
catch (Exception ex)
{
throw new InvalidOperationException(Resources.InvalidBAMFile, ex);
}
WriteHeader(header);
if (!HeaderOnly)
{
if (!string.IsNullOrEmpty(Library))
{
rgRecFields = header.RecordFields.Where(R => R.Typecode.ToUpper().Equals("RG")).ToList();
}
foreach (SAMAlignedSequence alignedSequence in GetAlignedSequence(stream))
{
WriteAlignedSequence(header, alignedSequence);
}
}
}
}
// Validates alignment header.
private static void ValidateAlignmentHeader(SAMAlignmentHeader header)
{
string message = header.IsValid();
if (!string.IsNullOrEmpty(message))
{
throw new ArgumentException(message);
}
}
/// <summary>
/// Parses the BAM file and returns the Header.
/// </summary>
private SAMAlignmentHeader GetHeader()
{
var header = new SAMAlignmentHeader();
RefSeqNames = new RegexValidatedStringList(SAMAlignedSequenceHeader.RNameRegxExprPattern);
_refSeqLengths = new List <int>();
ReadStream.Seek(0, SeekOrigin.Begin);
_deCompressedStream = null;
var array = new byte[8];
ReadUnCompressedData(array, 0, 8);
var lText = Helper.GetInt32(array, 4);
var samHeaderData = new byte[lText];
if (lText != 0)
{
ReadUnCompressedData(samHeaderData, 0, lText);
}
ReadUnCompressedData(array, 0, 4);
var noofRefSeqs = Helper.GetInt32(array, 0);
for (var i = 0; i < noofRefSeqs; i++)
{
ReadUnCompressedData(array, 0, 4);
var len = Helper.GetInt32(array, 0);
var refName = new byte[len];
ReadUnCompressedData(refName, 0, len);
ReadUnCompressedData(array, 0, 4);
var refLen = Helper.GetInt32(array, 0);
RefSeqNames.Add(Encoding.ASCII.GetString(refName, 0, refName.Length - 1));
_refSeqLengths.Add(refLen);
}
if (samHeaderData.Length != 0)
{
var str = Encoding.ASCII.GetString(samHeaderData);
using (var reader = new StringReader(str))
{
header = SAMParser.ParseSAMHeader(reader);
}
}
header.ReferenceSequences.Clear();
for (var i = 0; i < RefSeqNames.Count; i++)
{
var refname = RefSeqNames[i];
var length = _refSeqLengths[i];
header.ReferenceSequences.Add(new ReferenceSequenceInfo(refname, length));
}
return(header);
}
/// <summary>
/// Writes BAM header to the specified stream in BAM format.
/// </summary>
/// <param name="header">SAMAlignmentHeader object</param>
/// <param name="writer">Stream to write.</param>
public void WriteHeader(SAMAlignmentHeader header, Stream writer)
{
if (header == null)
{
throw new ArgumentNullException("header");
}
if (writer == null)
{
throw new ArgumentNullException("writer");
}
string samHeader;
if (_refSequences == null)
{
_refSequences = SortSequenceRanges(header.GetReferenceSequenceRanges());
}
using (StringWriter strwriter = new StringWriter(CultureInfo.InvariantCulture))
{
SAMFormatter.WriteHeader(header, strwriter);
samHeader = strwriter.ToString();
}
int samHeaderLen = samHeader.Length;
byte[] bytes = System.Text.ASCIIEncoding.ASCII.GetBytes(samHeader);
byte[] bamMagicNumber = new byte[4] {
66, 65, 77, 1
};
// write BAM magic number
writer.Write(bamMagicNumber, 0, 4);
// Length of the header text
writer.Write(Helper.GetLittleEndianByteArray(samHeaderLen), 0, 4);
//Plain header text in SAM
writer.Write(bytes, 0, bytes.Length);
// number of reference sequences
writer.Write(Helper.GetLittleEndianByteArray(_refSequences.Count), 0, 4);
for (int i = 0; i < _refSequences.Count; i++)
{
int len = _refSequences[i].ID.Length;
byte[] array = System.Text.ASCIIEncoding.ASCII.GetBytes(_refSequences[i].ID);
writer.Write(Helper.GetLittleEndianByteArray(len + 1), 0, 4);
writer.Write(array, 0, len);
writer.WriteByte((byte)'\0');
writer.Write(Helper.GetLittleEndianByteArray((int)_refSequences[i].End), 0, 4);
}
}
// Validates the alignment.
private SequenceAlignmentMap ValidateAlignment(ISequenceAlignment sequenceAlignment)
{
SequenceAlignmentMap seqAlignmentMap = sequenceAlignment as SequenceAlignmentMap;
if (seqAlignmentMap != null)
{
ValidateAlignmentHeader(seqAlignmentMap.Header);
if (CreateSortedBAMFile && SortType == BAMSortByFields.ChromosomeNameAndCoordinates)
{
this.refSequences = SortSequenceRanges(seqAlignmentMap.Header.GetReferenceSequenceRanges());
}
else
{
this.refSequences = seqAlignmentMap.Header.GetReferenceSequenceRanges();
}
return(seqAlignmentMap);
}
SAMAlignmentHeader header = sequenceAlignment.Metadata[Helper.SAMAlignmentHeaderKey] as SAMAlignmentHeader;
if (header == null)
{
throw new ArgumentException(Properties.Resource.SAMAlignmentHeaderNotFound);
}
ValidateAlignmentHeader(header);
seqAlignmentMap = new SequenceAlignmentMap(header);
if (CreateSortedBAMFile && SortType == BAMSortByFields.ChromosomeNameAndCoordinates)
{
this.refSequences = SortSequenceRanges(seqAlignmentMap.Header.GetReferenceSequenceRanges());
}
else
{
this.refSequences = seqAlignmentMap.Header.GetReferenceSequenceRanges();
}
foreach (IAlignedSequence alignedSeq in sequenceAlignment.AlignedSequences)
{
SAMAlignedSequenceHeader alignedHeader = alignedSeq.Metadata[Helper.SAMAlignedSequenceHeaderKey] as SAMAlignedSequenceHeader;
if (alignedHeader == null)
{
throw new ArgumentException(Properties.Resource.SAMAlignedSequenceHeaderNotFound);
}
SAMAlignedSequence samAlignedSeq = new SAMAlignedSequence(alignedHeader);
samAlignedSeq.QuerySequence = alignedSeq.Sequences[0];
seqAlignmentMap.QuerySequences.Add(samAlignedSeq);
}
return(seqAlignmentMap);
}
/// <summary>
/// Writes aligned sequence to output stream.
/// </summary>
/// <param name="header">Alignment header.</param>
/// <param name="alignedSequence">Aligned sequence to write.</param>
private void WriteAlignedSequence(SAMAlignmentHeader header, SAMAlignedSequence alignedSequence)
{
if (UnCompressedBAM || BAMOutput)
{
// Incase of compressed bamoutput uncompressed file will be compressed before sending it to output stream.
bamformatter.WriteAlignedSequence(header, alignedSequence, bamUncompressedOutStream);
}
else
{
SAMFormatter.WriteSAMAlignedSequence(alignedSequence, writer);
}
}
/// <summary>
/// Writes BAM header to the specified stream in BAM format.
/// </summary>
/// <param name="header">SAMAlignmentHeader object</param>
/// <param name="writer">Stream to write.</param>
public void WriteHeader(SAMAlignmentHeader header, Stream writer)
{
if (header == null)
{
throw new ArgumentNullException("header");
}
if (writer == null)
{
throw new ArgumentNullException("writer");
}
string samHeader;
if (this.refSequences == null)
{
this.refSequences = header.GetReferenceSequenceRanges();
}
using (StringWriter strwriter = new StringWriter(CultureInfo.InvariantCulture))
{
SAMFormatter.WriteHeader(strwriter, header);
samHeader = strwriter.ToString();
}
int samHeaderLen = samHeader.Length;
byte[] bytes = Encoding.UTF8.GetBytes(samHeader);
byte[] bamMagicNumber = { 66, 65, 77, 1 };
// write BAM magic number
writer.Write(bamMagicNumber, 0, 4);
// Length of the header text
writer.Write(Helper.GetLittleEndianByteArray(samHeaderLen), 0, 4);
//Plain header text in SAM
writer.Write(bytes, 0, bytes.Length);
// number of reference sequences
writer.Write(Helper.GetLittleEndianByteArray(this.refSequences.Count), 0, 4);
foreach (SequenceRange range in this.refSequences)
{
int len = range.ID.Length;
byte[] array = Encoding.UTF8.GetBytes(range.ID);
writer.Write(Helper.GetLittleEndianByteArray(len + 1), 0, 4);
writer.Write(array, 0, len);
writer.WriteByte((byte)'\0');
writer.Write(Helper.GetLittleEndianByteArray((int)range.End), 0, 4);
}
}
// Validates the alignment.
private SequenceAlignmentMap ValidateAlignment(ISequenceAlignment sequenceAlignment)
{
SequenceAlignmentMap seqAlignmentMap = sequenceAlignment as SequenceAlignmentMap;
if (seqAlignmentMap != null)
{
ValidateAlignmentHeader(seqAlignmentMap.Header);
_refSequences = SortSequenceRanges(seqAlignmentMap.Header.GetReferenceSequenceRanges());
foreach (SAMAlignedSequence alignedSequence in seqAlignmentMap.QuerySequences)
{
string message = alignedSequence.IsValidHeader();
if (!string.IsNullOrEmpty(message))
{
throw new ArgumentException(message);
}
ValidateSQHeader(alignedSequence.RName);
}
return(seqAlignmentMap);
}
SAMAlignmentHeader header = sequenceAlignment.Metadata[Helper.SAMAlignmentHeaderKey] as SAMAlignmentHeader;
if (header == null)
{
throw new ArgumentException(Resource.SAMAlignmentHeaderNotFound);
}
ValidateAlignmentHeader(header);
seqAlignmentMap = new SequenceAlignmentMap(header);
_refSequences = SortSequenceRanges(seqAlignmentMap.Header.GetReferenceSequenceRanges());
foreach (IAlignedSequence alignedSeq in sequenceAlignment.AlignedSequences)
{
SAMAlignedSequenceHeader alignedHeader = alignedSeq.Metadata[Helper.SAMAlignedSequenceHeaderKey] as SAMAlignedSequenceHeader;
if (alignedHeader == null)
{
throw new ArgumentException(Resource.SAMAlignedSequenceHeaderNotFound);
}
ValidateAlignedSequenceHeader(alignedHeader);
ValidateSQHeader(alignedHeader.RName);
SAMAlignedSequence samAlignedSeq = new SAMAlignedSequence(alignedHeader);
samAlignedSeq.QuerySequence = alignedSeq.Sequences[0];
}
return(seqAlignmentMap);
}
public void InvalidateSAMWriteTextWriter()
{
SAMAlignmentHeader header = new SAMAlignmentHeader();
try
{
SAMFormatter.WriteHeader(null, header);
Assert.Fail();
}
catch (ArgumentNullException)
{
ApplicationLog.WriteLine(
"SAM Formatter P2 : Successfully validated the exception");
}
}
/// <summary>
/// Writes the header to output stream
/// </summary>
/// <param name="header"></param>
private void WriteHeader(SAMAlignmentHeader header)
{
if (!Header && !HeaderOnly)
{
return;
}
if (UnCompressedBAM || BAMOutput)
{
// Incase of compressed bamoutput uncompressed file will be compressed before sending it to output stream.
bamformatter.WriteHeader(header, bamUncompressedOutStream);
}
else
{
SAMFormatter.WriteHeader(header, writer);
}
}
void validateInputFileAndLoadSampleNames()
{
if (!File.Exists(InputFilename))
{
throw new FileNotFoundException("Could not find file: " + InputFilename);
}
using (Stream stream = new FileStream(InputFilename, FileMode.Open, FileAccess.Read))
{
BAMParser bp = new BAMParser();
header = bp.GetHeader(stream);
var tmp = header.RecordFields.Where(x => x.Typecode == "RG").ToList();
sampleNames = tmp.Select(z => z.Tags.Where(p => p.Tag == "ID").First()).Select(z => z.Value).ToList();
NumberOfReadGroups = sampleNames.Count;
NumerOfSequences = header.ReferenceSequences.Count;
}
Console.WriteLine("Processing file with " + sampleNames.Count.ToString() + " samples and " + NumerOfSequences.ToString() + " reference sequences.");
}
/// <summary>
/// Writes SAMAlignedSequence to specified stream.
/// </summary>
/// <param name="header">Header from SAM object.</param>
/// <param name="alignedSeq">SAMAlignedSequence object.</param>
/// <param name="writer">Stream to write.</param>
public void WriteAlignedSequence(SAMAlignmentHeader header, SAMAlignedSequence alignedSeq, Stream writer)
{
if (header == null)
{
throw new ArgumentNullException("header");
}
if (alignedSeq == null)
{
throw new ArgumentNullException("alignedSeq");
}
if (writer == null)
{
throw new ArgumentNullException("writer");
}
_refSequences = SortSequenceRanges(header.GetReferenceSequenceRanges());
WriteAlignedSequence(alignedSeq, writer);
}
public IEnumerable <CompactSAMSequence> ParseRangeAsEnumerableSequences(string fileName, string refSeqName, int start = 0, int end = Int32.MaxValue)
{
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))
{
readStream = bamStream;
if (readStream == null || readStream.Length == 0)
{
throw new FileFormatException(Properties.Resource.BAM_InvalidBAMFile);
}
ValidateReader();
SAMAlignmentHeader header = GetHeader();
// verify whether there is any reads related to chromosome.
int refSeqIndex = refSeqNames.IndexOf(refSeqName);
if (refSeqIndex < 0)
{
string message = string.Format(CultureInfo.InvariantCulture, Properties.Resource.BAM_RefSeqNotFound, refSeqName);
throw new ArgumentException(message, "refSeqName");
}
BAMIndex bamIndexInfo = bamIndexFile.Read();
BAMReferenceIndexes refIndex = bamIndexInfo.RefIndexes[refSeqIndex];
IList <Chunk> chunks = GetChunks(refIndex, start, end);
foreach (var s in EnumerateAlignedSequences(chunks))
{
if (s != null && (s.RName == "*" || (s.Pos >= (start - 1) && s.RefEndPos < end)))
{
yield return(s);
}
}
readStream = null;
}
}
}
// Validates the alignment.
private SequenceAlignmentMap ValidateAlignment(ISequenceAlignment sequenceAlignment)
{
SequenceAlignmentMap seqAlignmentMap = sequenceAlignment as SequenceAlignmentMap;
if (seqAlignmentMap != null)
{
ValidateAlignmentHeader(seqAlignmentMap.Header);
_refSequences = SortSequenceRanges(seqAlignmentMap.Header.GetReferenceSequenceRanges());
return(seqAlignmentMap);
}
SAMAlignmentHeader header = sequenceAlignment.Metadata[Helper.SAMAlignmentHeaderKey] as SAMAlignmentHeader;
if (header == null)
{
throw new ArgumentException(Resource.SAMAlignmentHeaderNotFound);
}
ValidateAlignmentHeader(header);
seqAlignmentMap = new SequenceAlignmentMap(header);
_refSequences = SortSequenceRanges(seqAlignmentMap.Header.GetReferenceSequenceRanges());
foreach (IAlignedSequence alignedSeq in sequenceAlignment.AlignedSequences)
{
SAMAlignedSequenceHeader alignedHeader = alignedSeq.Metadata[Helper.SAMAlignedSequenceHeaderKey] as SAMAlignedSequenceHeader;
if (alignedHeader == null)
{
throw new ArgumentException(Resource.SAMAlignedSequenceHeaderNotFound);
}
SAMAlignedSequence samAlignedSeq = new SAMAlignedSequence(alignedHeader);
samAlignedSeq.QuerySequence = alignedSeq.Sequences[0];
seqAlignmentMap.QuerySequences.Add(samAlignedSeq);
}
return(seqAlignmentMap);
}
/// <summary>
/// Updates the header with reference name from reads in input file.
/// </summary>
/// <param name="header">SAM alignment header.</param>
private void UpdateReferenceInformationFromReads(SAMAlignmentHeader header)
{
// If the ReferenceNamesAndLength file name is not specified and there is no @SQ header,
// then get the refernece names from read information.
List <string> refSeqNames = new List <string>();
using (StreamReader textReader = new StreamReader(InputFilePath))
{
foreach (SAMAlignedSequence alignedSeq in GetAlignedSequence(textReader))
{
if (!alignedSeq.RName.Equals("*", StringComparison.OrdinalIgnoreCase) &&
!refSeqNames.Contains(alignedSeq.RName, StringComparer.OrdinalIgnoreCase))
{
refSeqNames.Add(alignedSeq.RName);
}
}
}
foreach (string refname in refSeqNames)
{
header.ReferenceSequences.Add(new ReferenceSequenceInfo(refname, 0));
}
}
/// <summary>
/// Comapare Sequence Alignment Header fields
/// </summary>
/// <param name="actualAlignment">Actual sequence alignment object</param>
/// <param name="expectedAlignment">Expected sequence alignment object</param>
/// <returns></returns>
bool CompareSequencedAlignmentHeader(SequenceAlignmentMap actualAlignment,
SequenceAlignmentMap expectedAlignment)
{
SAMAlignmentHeader aheader = actualAlignment.Header;
IList <SAMRecordField> arecordFields = aheader.RecordFields;
SAMAlignmentHeader expectedheader = expectedAlignment.Header;
IList <SAMRecordField> expectedrecordFields = expectedheader.RecordFields;
int tagKeysCount = 0;
int tagValuesCount = 0;
for (int index = 0; index < expectedrecordFields.Count; index++)
{
if (0 != string.Compare(expectedrecordFields[index].Typecode.ToString(),
arecordFields[index].Typecode.ToString()))
{
return(false);
}
for (int tags = 0; tags < expectedrecordFields[index].Tags.Count; tags++)
{
if ((0 != string.Compare(expectedrecordFields[index].Tags[tags].Tag.ToString(),
arecordFields[index].Tags[tags].Tag.ToString())) ||
(0 != string.Compare(expectedrecordFields[index].Tags[tags].Value.ToString(),
arecordFields[index].Tags[tags].Value.ToString())))
{
Console.WriteLine(string.Format(null,
"SAM Parser BVT : Sequence alignment header does not match"));
ApplicationLog.WriteLine(string.Format(null,
"SAM Parser BVT : Sequence alignment header does not match"));
return(false);
}
tagKeysCount++;
tagValuesCount++;
}
}
return(true);
}
/// <summary>
/// Merge multiple sorted alignments.
/// SAMUtil.exe out.bam in1.bam in2.bam
/// </summary>
public void DoMerge()
{
if (FilePaths == null)
{
throw new InvalidOperationException("FilePath");
}
if (FilePaths.Length < 3)
{
throw new InvalidOperationException(Resources.MergeHelp);
}
IList <IList <BAMSortedIndex> > sortedIndexes = new List <IList <BAMSortedIndex> >();
IList <SequenceAlignmentMap> sequenceAlignmentMaps = new List <SequenceAlignmentMap>();
IList <int> help = new List <int>();
Parallel.For(1, FilePaths.Length, (int index) =>
{
IList <BAMSortedIndex> sortedIndex;
BAMParser parser = new BAMParser();;
SequenceAlignmentMap map;
if (index == 1)
{
try
{
map = parser.Parse(FilePaths[1]);
}
catch
{
throw new InvalidOperationException(Resources.InvalidBAMFile);
}
if (map == null)
{
throw new InvalidOperationException(Resources.EmptyFile);
}
if (string.IsNullOrEmpty(HeaderFile) && map.Header.RecordFields.Count == 0)
{
throw new InvalidOperationException(Resources.HeaderMissing);
}
if (!string.IsNullOrEmpty(HeaderFile))
{
SAMParser parse = new SAMParser();
SequenceAlignmentMap head;
try
{
head = parse.Parse(HeaderFile);
}
catch
{
throw new InvalidOperationException(Resources.IncorrectHeaderFile);
}
if (head == null)
{
throw new InvalidOperationException(Resources.EmptyFile);
}
_header = head.Header;
}
else
{
_header = map.Header;
}
sortedIndex = Sort(map, SortByReadName ? BAMSortByFields.ReadNames : BAMSortByFields.ChromosomeCoordinates);
}
else
{
try
{
map = parser.Parse(FilePaths[index]);
}
catch
{
throw new InvalidOperationException(Resources.InvalidBAMFile);
}
if (map == null)
{
throw new InvalidOperationException(Resources.EmptyFile);
}
sortedIndex = Sort(map, SortByReadName ? BAMSortByFields.ReadNames : BAMSortByFields.ChromosomeCoordinates);
}
lock (sortedIndexes)
{
sortedIndexes.Add(sortedIndex);
sequenceAlignmentMaps.Add(map);
}
});
string filePath = Path.GetTempFileName();
using (FileStream fstemp = new FileStream(filePath, FileMode.Create, FileAccess.ReadWrite))
{
BAMFormatter formatter = new BAMFormatter();
formatter.WriteHeader(_header, fstemp);
int[] indexes = new int[sortedIndexes.Count];
if (SortByReadName)
{
IList <BAMSortedIndex> sortedIndex = sortedIndexes.Select(a => a.First()).ToList();
WriteMergeFileSortedByReadName(sortedIndex, fstemp, formatter, sequenceAlignmentMaps);
}
else
{
WriteMergeFile(sortedIndexes, fstemp, formatter, sequenceAlignmentMaps);
}
using (FileStream fsoutput = new FileStream(FilePaths[0], FileMode.Create, FileAccess.Write))
{
fstemp.Seek(0, SeekOrigin.Begin);
formatter.CompressBAMFile(fstemp, fsoutput);
}
}
File.Delete(filePath);
}
// Gets new header with sorted SQ Fields.
// If SQ fields are already sorted then returns the same header.
private SAMAlignmentHeader GetHeaderWithSortedSQFields(SAMAlignmentHeader header, bool canChangeOtherTagPos)
{
if (IsSortedByChromosomeNames(GetSQHeaders(header.RecordFields)))
{
return(header);
}
SAMAlignmentHeader newHeader = new SAMAlignmentHeader();
int i = 0;
if (canChangeOtherTagPos)
{
List <SAMRecordField> sqHeaders = new List <SAMRecordField>();
for (; i < header.RecordFields.Count; i++)
{
SAMRecordField field = header.RecordFields[i];
if (field.Typecode.Equals("SQ"))
{
sqHeaders.Add(field);
}
else
{
newHeader.RecordFields.Add(field);
}
sqHeaders.Sort(CompareByChromosomeName);
foreach (SAMRecordField sqfield in sqHeaders)
{
newHeader.RecordFields.Add(sqfield);
}
foreach (string str in header.Comments)
{
newHeader.Comments.Add(str);
}
}
}
else
{
Bio.Util.SortedList <SAMRecordField, int> map = new Bio.Util.SortedList <SAMRecordField, int>(new ComparisonWrapper <SAMRecordField>(CompareByChromosomeName));
for (; i < header.RecordFields.Count; i++)
{
SAMRecordField field = header.RecordFields[i];
if (field.Typecode.Equals("SQ"))
{
map.Add(field, i);
}
newHeader.RecordFields.Add(field);
}
i = 0;
foreach (int index in map.Values.OrderBy(I => I))
{
newHeader.RecordFields[index] = map.Keys[i++];
}
foreach (string str in header.Comments)
{
newHeader.Comments.Add(str);
}
}
return(newHeader);
}
/// <summary>
/// Run ConPADE on each contig of the input BAM file.
/// </summary>
/// <param name="bamName">Name of the input BAM file.</param>
public void RunFile(string bamName)
{
// Current implementation requires that minimum ploidy be 1
int min_ploidy = 1;
int number_of_ploidies = max_ploidy - min_ploidy + 1;
// Set nucleotide proportions (genotypes)
double[][][] nuc_props = Nuc_Props(min_ploidy, number_of_ploidies);
// Set dosage probabilities
double SNP_density = (double)1 / snpDens;
double no_SNP_prob = Math.Log((1 - SNP_density) / 2);
double[][] dose_probs = Dose_Probs(min_ploidy, number_of_ploidies, SNP_density, no_SNP_prob);
// Set HiSeq error model
double[, , , ,] log_probs = Error_Probs();
// Set substitution model
double[, ,] log_subst_probs = Subst_Probs();
// Set SNP calling probability
double log_SNP_thres = SNPthres * Math.Log(10) / -10;
Stopwatch clock = new Stopwatch();
Console.WriteLine("Program started at {0}\n", DateTime.Now);
Stream bam_stream = new FileStream(bamName, FileMode.Open, FileAccess.Read);
BAMParser parser = new BAMParser();
SAMAlignmentHeader header = parser.GetHeader(bam_stream);
string temp = Path.GetFileNameWithoutExtension(bamName);
// Find first valid alignment in BAM file
SAMAlignedSequence next_alignment = parser.GetAlignedSequence(true);
while (next_alignment == null || next_alignment.RName == "*" || next_alignment.IsDummyRead)
{
next_alignment = parser.GetAlignedSequence(true);
}
TextWriter writer_log_like = null;
TextWriter writer_SNP = null;
TextWriter writer_ploidy = null;
TextWriter writer_reads = null;
// Create global output files and write headers.
if (!splitContigs)
{
string SNP_file = temp + "_SNP.txt";
writer_SNP = new StreamWriter(SNP_file);
writer_SNP.WriteLine("Contig\tPosition\tAlleles\tCounts\tDosage\tPhredQuality");
string ploidy_file = temp + "_ploidy.txt";
writer_ploidy = new StreamWriter(ploidy_file);
writer_ploidy.Write("Contig\tBestPloidy");
for (int i = 0; i < number_of_ploidies; i++)
{
writer_ploidy.Write("\tlogLike_M{0}", i + min_ploidy);
}
writer_ploidy.WriteLine("");
string reads_file = temp + "_readStats.txt";
writer_reads = new StreamWriter(reads_file);
writer_reads.WriteLine("Contig\tAlignedReads\tAlignedBases\tUsedReads\tUsedBases");
}
// Run over each contig in input BAM file.
int contig_ind = -1;
while (next_alignment != null && next_alignment.RName != "*" && !next_alignment.IsDummyRead)
{
string contig_name = next_alignment.RName;
Console.WriteLine("Started contig {0} at {1}",
contig_name, DateTime.Now);
clock.Restart();
#region Variables and file handles for current contig
long number_of_aligned_reads = 0;
long number_of_aligned_base_pairs = 0;
long number_of_used_reads = 0;
long number_of_used_base_pairs = 0;
// Create individual output files for the current contig.
if (splitContigs)
{
string name = temp + "_" + contig_name;
string log_like_file = name + "_log_likelihoods.txt";
writer_log_like = new StreamWriter(log_like_file);
string SNP_file = name + "_SNP.txt";
writer_SNP = new StreamWriter(SNP_file);
string ploidy_file = name + "_ploidy.txt";
writer_ploidy = new StreamWriter(ploidy_file);
string reads_file = name + "_readStats.txt";
writer_reads = new StreamWriter(reads_file);
}
double[] global_log_like = new double[number_of_ploidies];
while (header.ReferenceSequences[++contig_ind].Name != contig_name)
{
;
}
long contig_length = header.ReferenceSequences[contig_ind].Length;
// Create a queue to include all reads that overlap with a given position.
Queue <Padded_Read> read_queue = new Queue <Padded_Read>();
// Create a queue to include best doses for each tested position.
Queue <Best_Dose> dose_queue = new Queue <Best_Dose>((int)contig_length);
#endregion Variables and file handles for current contig
int positions_to_compute = 0;
long current_position = 0;
#region Run over every position in contig
while (current_position < contig_length)
{
if ((current_position % 1000000) == 0 && current_position != 0)
{
Console.WriteLine("At position {0} of {1}", current_position + 1, contig_length);
}
// Search for reads starting at current position.
Search_Reads(parser, ref next_alignment, contig_name, ref number_of_aligned_reads,
ref number_of_aligned_base_pairs, ref number_of_used_reads, ref number_of_used_base_pairs,
read_queue, current_position);
if (read_queue.Count > 0)
{
positions_to_compute++;
// Extract information from each read in queue.
byte[] obs_nucs;
byte[] is_GG;
bool[] reverse;
int[] quality_scores;
int[] neigh_quality_scores;
int[] scores;
int[] counts;
int k;
Extract_Read_Info(read_queue, current_position, out obs_nucs, out is_GG, out reverse,
out quality_scores, out neigh_quality_scores, out scores, out counts, out k);
// Find two most abundant nucleotides for this position.
byte nuc_one;
byte nuc_two;
Get_Two_Nucs(scores, out nuc_one, out nuc_two);
// Calculate Pr(obs|allele1) and Pr(obs|allele2).
double[][] log_nuc_probs = Obs_Probs(log_probs, log_subst_probs, obs_nucs, is_GG, reverse,
quality_scores, neigh_quality_scores, counts, k, nuc_one, nuc_two);
// Calculate log_likelihoods of genotypes for current position.
double[][] log_likelihoods = Log_Likelihoods(min_ploidy, max_ploidy, log_nuc_probs, nuc_props);
// Calculate log_likelihood of each ploidy and keep most likely allele dosage.
Global_Likelihood_Keep_Dose(min_ploidy, number_of_ploidies, dose_probs, global_log_like,
dose_queue, current_position, counts, nuc_one, nuc_two, log_likelihoods);
}
// Remove finished reads from queue. Finished reads no longer overlap with current position.
Padded_Read read_to_remove;
if (read_queue.Count > 0)
{
read_to_remove = read_queue.First();
}
else
{
read_to_remove = null;
}
while (read_to_remove != null &&
(read_to_remove.alignment.Pos + read_to_remove.alignment_length - 2) < current_position)
{
read_queue.Dequeue();
if (read_queue.Count > 0)
{
read_to_remove = read_queue.First();
}
else
{
read_to_remove = null;
}
}
++current_position;
}
#endregion Run over every position in contig
// Output log_likelihoods.
int best_log_like = 0;
for (int i = 0; i < number_of_ploidies; i++)
{
if (global_log_like[i] > global_log_like[best_log_like])
{
best_log_like = i;
}
if (splitContigs)
{
writer_log_like.WriteLine("Ploidy {0} - log_likelihood {1}", i + min_ploidy, global_log_like[i]);
}
}
// Output most likely ploidy.
int best_ploidy = best_log_like + min_ploidy;
if (splitContigs)
{
writer_ploidy.WriteLine(best_ploidy);
}
else
{
writer_ploidy.Write("{0}\t{1}", contig_name, best_ploidy);
for (int i = 0; i < number_of_ploidies; i++)
{
writer_ploidy.Write("\t{0}", global_log_like[i]);
}
writer_ploidy.WriteLine("");
}
// Output SNPs.
if (splitContigs)
{
writer_SNP.WriteLine("Position\tAlleles\tCounts\tDosage\tPhredQuality");
}
char[] nuc_chars = new char[4] {
'A', 'C', 'G', 'T'
};
foreach (Best_Dose cur_doses in dose_queue)
{
double cur_SNP_posterior = cur_doses.SNP_posterior[best_log_like];
if (cur_SNP_posterior <= log_SNP_thres)
{
int cur_best_dose = cur_doses.best_dose[best_log_like];
if (cur_best_dose != best_ploidy && cur_best_dose != 0)
{
if (splitContigs)
{
writer_SNP.WriteLine("{0}\t{1}|{2}\t{3}|{4}\t{5}\t{6}", cur_doses.position + 1,
nuc_chars[cur_doses.nuc_one], nuc_chars[cur_doses.nuc_two], cur_doses.count_one,
cur_doses.count_two, cur_best_dose, -10 * cur_SNP_posterior / Math.Log(10));
}
else
{
writer_SNP.WriteLine("{0}\t{1}\t{2}|{3}\t{4}|{5}\t{6}\t{7}", contig_name,
cur_doses.position + 1, nuc_chars[cur_doses.nuc_one], nuc_chars[cur_doses.nuc_two],
cur_doses.count_one, cur_doses.count_two, cur_best_dose,
-10 * cur_SNP_posterior / Math.Log(10));
}
}
}
}
// Output read statistics.
if (splitContigs)
{
writer_reads.WriteLine("\nNumber of aligned reads: {0}", number_of_aligned_reads);
writer_reads.WriteLine("Number of aligned base pairs: {0}", number_of_aligned_base_pairs);
writer_reads.WriteLine("\nNumber of used reads: {0}", number_of_used_reads);
writer_reads.WriteLine("Number of used base pairs: {0}", number_of_used_base_pairs);
}
else
{
writer_reads.WriteLine("{0}\t{1}\t{2}\t{3}\t{4}", contig_name, number_of_aligned_reads,
number_of_aligned_base_pairs, number_of_used_reads, number_of_used_base_pairs);
}
if (splitContigs)
{
writer_log_like.Close();
writer_SNP.Close();
writer_ploidy.Close();
writer_reads.Close();
}
clock.Stop();
Console.WriteLine("Time to run contig: {0} s\n", (double)clock.ElapsedMilliseconds / 1000);
}
if (!splitContigs)
{
writer_SNP.Close();
writer_ploidy.Close();
writer_reads.Close();
}
parser.Dispose();
Console.WriteLine("Finished at {0}\n", DateTime.Now);
}
public IEnumerable <CompactSAMSequence> Parse()
{
if (string.IsNullOrWhiteSpace(_fileName))
{
throw new ArgumentNullException("fileName");
}
using (readStream = new FileStream(_fileName, FileMode.Open, FileAccess.Read, FileShare.Read))
{
Stream reader = readStream;
if (reader == null || reader.Length == 0)
{
throw new FileFormatException(Properties.Resource.BAM_InvalidBAMFile);
}
if (!String.IsNullOrEmpty(ChromosomeToGet))
{
foreach (var s in
ParseRangeAsEnumerableSequences(_fileName, ChromosomeToGet))
{
if (s != null)
{
yield return(s);
}
////TODO: Super inefficient right now, am parsing the sequence multiple times,
////fix this.
//var s2 = s.ToArray ();
//var alpha = Alphabets.AutoDetectAlphabet(s2, 0, s2.Length, null);
//var strippedOfInfo = new Sequence(alpha, s2);
//yield return strippedOfInfo;
}
}
else
{
readStream = reader;
ValidateReader();
SAMAlignmentHeader header = GetHeader();
SequenceAlignmentMap sequenceAlignmentMap = null;
if (sequenceAlignmentMap == null)
{
sequenceAlignmentMap = new SequenceAlignmentMap(header);
}
while (!IsEOF())
{
#if WANT_OLD_VERSION
SAMAlignedSequence alignedSeq = GetAlignedSequence(0, int.MaxValue);
#else
var alignedSeq = GetAlignedSequence();
#endif
if (alignedSeq != null)
{
#if WANT_OLD_VERSION
//make a new Sequence
ISequence strippedOfInfo = null;
try
{
var syms = alignedSeq.QuerySequence.ToArray();
var alpha = Alphabets.AutoDetectAlphabet(syms, 0, syms.Length, null);
strippedOfInfo = new Sequence(alpha, alignedSeq.QuerySequence.ToArray());
strippedOfInfo = alignedSeq;
}
catch (ArgumentOutOfRangeException exception)
{
Debug.Write("Could not convert sequence: " + exception.Message);
}
if (strippedOfInfo != null)
{
yield return(strippedOfInfo);
}
#else
yield return(alignedSeq);
#endif
}
alignedSeq = null;
}
}
}
}
/// <summary>
/// Converts the input SAM to BAM file format.
/// </summary>
private void ConvertFromSAMTOBAM()
{
SAMAlignmentHeader header = null;
try
{
header = SAMParser.ParseSAMHeader(InputFilePath);
}
catch (Exception ex)
{
throw new InvalidOperationException(Resources.InvalidSAMFile, ex);
}
if (header == null)
{
Console.Error.WriteLine("Warning: SAM file doesn't contian header");
}
if (HeaderOnly)
{
if (header != null)
{
WriteHeader(header);
}
}
else
{
if (header == null)
{
header = new SAMAlignmentHeader();
}
if (!string.IsNullOrEmpty(Library))
{
rgRecFields = header.RecordFields.Where(R => R.Typecode.ToUpper().Equals("RG")).ToList();
}
if (!string.IsNullOrEmpty(ReferenceNamesAndLength))
{
this.UpdateReferenceInformationFromFile(header);
}
else if (header.ReferenceSequences.Count == 0)
{
this.UpdateReferenceInformationFromReads(header);
}
WriteHeader(header);
using (StreamReader textReader = new StreamReader(InputFilePath))
{
foreach (SAMAlignedSequence alignedSeq in GetAlignedSequence(textReader))
{
WriteAlignedSequence(header, alignedSeq);
}
}
}
if (UnCompressedBAM)
{
bamUncompressedOutStream.Flush();
if (writer != null)
{
DisplayBAMContent(bamUncompressedOutStream);
}
}
if (BAMOutput && !UnCompressedBAM)
{
bamUncompressedOutStream.Flush();
bamUncompressedOutStream.Seek(0, SeekOrigin.Begin);
bamformatter.CompressBAMFile(bamUncompressedOutStream, bamCompressedOutStream);
bamCompressedOutStream.Flush();
if (writer != null)
{
DisplayBAMContent(bamCompressedOutStream);
}
}
}
/// <summary>
/// Extract/print all or sub alignments in SAM or BAM format.
/// By default, this command assumes the file on the command line is in
/// BAM format and it prints the alignments in SAM.
/// SAMUtil.exe view in.bam
/// </summary>
public void ViewResult()
{
try
{
if (string.IsNullOrEmpty(InputFilePath))
{
throw new InvalidOperationException("Input File Not specified");
}
if (!string.IsNullOrEmpty(Region))
{
StringToRegionConverter();
}
Initialize();
SAMAlignmentHeader header = null;
if (!SAMInput)
{
Stream stream = new FileStream(InputFilePath, FileMode.Open, FileAccess.Read);
try
{
header = bamparser.GetHeader(stream);
}
catch
{
throw new InvalidOperationException(Resources.InvalidBAMFile);
}
WriteHeader(header);
if (!HeaderOnly)
{
if (!string.IsNullOrEmpty(Library))
{
rgRecFields = header.RecordFields.Where(R => R.Typecode.ToUpper().Equals("RG")).ToList();
}
foreach (SAMAlignedSequence alignedSequence in GetAlignedSequence(stream))
{
WriteAlignedSequence(header, alignedSequence);
}
}
}
else
{
try
{
header = SAMParser.ParseSAMHeader(InputFilePath);
}
catch
{
throw new InvalidOperationException(Resources.InvalidSAMFile);
}
if (header == null)
{
throw new InvalidOperationException("SAM file doesn't contian header");
}
WriteHeader(header);
if (!HeaderOnly)
{
if (!string.IsNullOrEmpty(Library))
{
rgRecFields = header.RecordFields.Where(R => R.Typecode.ToUpper().Equals("RG")).ToList();
}
using (StreamReader textReader = new StreamReader(InputFilePath))
{
foreach (SAMAlignedSequence alignedSeq in GetAlignedSequence(textReader))
{
WriteAlignedSequence(header, alignedSeq);
}
}
}
if (UnCompressedBAM)
{
bamUncompressedOutStream.Flush();
if (writer != null)
{
DisplayBAMContent(bamUncompressedOutStream);
}
}
if (BAMOutput && !UnCompressedBAM)
{
bamUncompressedOutStream.Flush();
bamUncompressedOutStream.Seek(0, SeekOrigin.Begin);
bamformatter.CompressBAMFile(bamUncompressedOutStream, bamCompressedOutStream);
bamCompressedOutStream.Flush();
if (writer != null)
{
DisplayBAMContent(bamCompressedOutStream);
}
}
}
}
finally
{
Close();
}
}