public void MapContigToReverseComplementOfRead()
{
IList <ISequence> contigs = new List <ISequence>();
IList <ISequence> reads = new List <ISequence>();
Sequence seq = new Sequence(Alphabets.DNA, "TCTGATAAGG".Select(a => (byte)a).ToArray());
seq.ID = "1";
contigs.Add(seq);
Sequence read = new Sequence(Alphabets.DNA, "CCTTATCAG".Select(a => (byte)a).ToArray());
read.ID = "2";
reads.Add(read);
const int kmerLength = 6;
IList <Contig> alignment = ReadAlignment.ReadContigAlignment(contigs, reads, kmerLength);
Assert.AreEqual(alignment.Count, contigs.Count);
Contig contig = alignment.First();
Contig.AssembledSequence sequence = contig.Sequences.First();
Assert.AreEqual(sequence.Length, 9);
Assert.AreEqual(sequence.Position, 1);
Assert.AreEqual(sequence.ReadPosition, 0);
Assert.AreEqual(sequence.Sequence, reads.First());
Assert.AreEqual(sequence.IsComplemented, true);
Assert.AreEqual(sequence.IsReversed, true);
}
public void MapReadToContig()
{
IList <ISequence> contigs = new List <ISequence>();
IList <ISequence> reads = new List <ISequence>();
Sequence seq = new Sequence(Alphabets.DNA, "TCTGATAAGG");
seq.DisplayID = "1";
contigs.Add(seq);
Sequence read = new Sequence(Alphabets.DNA, "CTGATAAGG");
read.DisplayID = "2";
reads.Add(read);
const int kmerLength = 6;
IList <Contig> alignment = ReadAlignment.ReadContigAlignment(contigs, reads, kmerLength);
Assert.AreEqual(alignment.Count, contigs.Count);
Contig contig = alignment.First();
Contig.AssembledSequence sequence = contig.Sequences.First();
Assert.AreEqual(sequence.Length, 9);
Assert.AreEqual(sequence.Position, 1);
Assert.AreEqual(sequence.ReadPosition, 0);
Assert.AreEqual(sequence.Sequence, reads.First());
Assert.AreEqual(sequence.IsComplemented, false);
Assert.AreEqual(sequence.IsReversed, false);
}
/// <summary>
/// Parses a list of assembled sparse sequences from the reader.
/// </summary>
/// <param name="contigReader">The reader to read the assembled sparse sequences from
/// Flag to indicate whether the resulting sequences should be in readonly mode or not.
/// If this flag is set to true then the resulting sequences's isReadOnly property
/// will be set to true, otherwise it will be set to false.
/// </param>
/// <returns>Returns contig assemble sequence.</returns>
protected IList <Contig.AssembledSequence> ParseAssembledSequence(XsvSparseReader contigReader)
{
// Check input arguments
if (contigReader == null)
{
throw new ArgumentNullException("contigReader");
}
List <Contig.AssembledSequence> sequenceList = new List <Contig.AssembledSequence>();
while (contigReader.HasLines)
{
Contig.AssembledSequence aseq = new Contig.AssembledSequence();
aseq.Sequence = ParseOne(contigReader);
sequenceList.Add(aseq);
}
return(sequenceList);
}
/// <summary>
/// Parses a list of assembled sparse sequences from the reader.
/// </summary>
/// <param name="contigReader">The reader to read the assembled sparse sequences from</param>
/// <param name="isReadOnly">
/// Flag to indicate whether the resulting sequences should be in readonly mode or not.
/// If this flag is set to true then the resulting sequences's isReadOnly property
/// will be set to true, otherwise it will be set to false.
/// </param>
/// <returns></returns>
protected IList <Contig.AssembledSequence> ParseAssembledSequence(XsvSparseReader contigReader, bool isReadOnly)
{
if (contigReader == null)
{
throw new ArgumentNullException("contigReader");
}
List <Contig.AssembledSequence> sequenceList = new List <Contig.AssembledSequence>();
while (contigReader.HasLines)
{
Contig.AssembledSequence aseq = new Contig.AssembledSequence();
int offset;
var sequenceWithOffset = ParseOneWithOffset(contigReader, isReadOnly);
aseq.Sequence = sequenceWithOffset.Item1;
offset = sequenceWithOffset.Item2;
aseq.Position = offset;
sequenceList.Add(aseq);
}
return(sequenceList);
}
/// <summary>
/// Aligns reads to contigs using kmer method of alignment.
/// </summary>
/// <param name="contigs">List of contig sequences.</param>
/// <param name="reads">List of read sequences.</param>
/// <param name="kmerLength">Kmer Length.</param>
/// <returns>List of Contig.</returns>
public static IList <Contig> ReadContigAlignment(IList <ISequence> contigs, IList <ISequence> reads, int kmerLength)
{
KmerIndexerDictionary map = SequenceToKmerBuilder.BuildKmerDictionary(reads, kmerLength);
IList <ContigIndex> contigDatas;
contigDatas = contigs.AsParallel().Select(contig =>
{
IEnumerable <ISequence> kmers = SequenceToKmerBuilder.GetKmerSequences(contig, kmerLength);
ContigIndex index = new ContigIndex(contig);
IList <KmerIndexer> positions;
foreach (ISequence kmer in kmers)
{
if (map.TryGetValue(kmer, out positions) ||
map.TryGetValue(kmer.GetReverseComplementedSequence(), out positions))
{
index.ContigReadMatchIndexes.Add(positions);
}
else
{
index.ContigReadMatchIndexes.Add(new List <KmerIndexer>());
}
}
return(index);
}).ToList();
return(contigDatas.Select(contigData =>
{
IList <Task <IList <ReadMap> > > tasks =
new List <Task <IList <ReadMap> > >();
// Stores information about contigs for which tasks has been generated.
IList <long> visitedReads = new List <long>();
// Creates Task for every read in nodes for a given contig.
for (int index = 0; index < contigData.ContigReadMatchIndexes.Count; index++)
{
int readPosition = index;
foreach (KmerIndexer kmer in contigData.ContigReadMatchIndexes[index])
{
long contigIndex = kmer.SequenceIndex;
if (!visitedReads.Contains(contigIndex))
{
visitedReads.Add(contigIndex);
tasks.Add(
Task <IList <ReadMap> > .Factory.StartNew(t => MapRead(readPosition, contigData.ContigReadMatchIndexes, contigIndex, kmerLength), TaskCreationOptions.AttachedToParent));
}
}
}
Contig contigOutputStructure = new Contig();
contigOutputStructure.Consensus = contigData.ContigSequence;
for (int index = 0; index < visitedReads.Count; index++)
{
foreach (ReadMap maps in tasks[index].Result)
{
Contig.AssembledSequence assembledSeq = new Contig.AssembledSequence()
{
Length = maps.Length,
Position = maps.StartPositionOfContig,
ReadPosition = maps.StartPositionOfRead,
Sequence = reads.ElementAt(visitedReads[index])
};
if (new string(
contigOutputStructure.Consensus.GetSubSequence(
assembledSeq.Position, assembledSeq.Length).Select(a => (char)a).ToArray()).
Equals(new string(assembledSeq.Sequence.GetSubSequence(assembledSeq.ReadPosition, assembledSeq.Length)
.Select(a => (char)a).ToArray())))
{
assembledSeq.IsComplemented = false;
assembledSeq.IsReversed = false;
}
else
{
assembledSeq.IsComplemented = true;
assembledSeq.IsReversed = true;
}
contigOutputStructure.Sequences.Add(assembledSeq);
}
}
return contigOutputStructure;
}).ToList());
}
public void TestContigWithBinaryFormatter()
{
Stream stream = null;
try
{
stream = File.Open("Contig.data", FileMode.Create);
BinaryFormatter formatter = new BinaryFormatter();
Sequence seq1 = new Sequence(Alphabets.DNA, "ACGACTTACG");
Contig.AssembledSequence assembledSeq1 = new Contig.AssembledSequence();
assembledSeq1.Sequence = seq1;
assembledSeq1.Position = 0;
assembledSeq1.IsReversed = false;
assembledSeq1.IsComplemented = false;
Sequence seq2 = new Sequence(Alphabets.DNA, "TACGATCCGGAAA");
Contig.AssembledSequence assembledSeq2 = new Contig.AssembledSequence();
assembledSeq2.Sequence = seq2;
assembledSeq2.Position = 6;
assembledSeq2.IsReversed = false;
assembledSeq2.IsComplemented = false;
Sequence consensus = new Sequence(Alphabets.DNA, "ACGACTTACGATCCGGAAA");
Contig contig = new Contig();
contig.Sequences.Add(assembledSeq1);
contig.Sequences.Add(assembledSeq2);
contig.Consensus = consensus;
formatter.Serialize(stream, contig);
stream.Seek(0, SeekOrigin.Begin);
Contig deserializedContig = (Contig)formatter.Deserialize(stream);
Assert.AreNotSame(contig, deserializedContig);
Assert.AreEqual(contig.Consensus.ToString(), deserializedContig.Consensus.ToString());
Assert.AreEqual(contig.Length, deserializedContig.Length);
Assert.AreEqual(contig.Sequences.Count, deserializedContig.Sequences.Count);
for (int i = 0; i < contig.Sequences.Count; i++)
{
Assert.AreEqual(
contig.Sequences[i].Sequence.ToString(),
deserializedContig.Sequences[i].Sequence.ToString());
Assert.AreEqual(
contig.Sequences[i].IsComplemented,
deserializedContig.Sequences[i].IsComplemented);
Assert.AreEqual(contig.Sequences[i].IsReversed, deserializedContig.Sequences[i].IsReversed);
Assert.AreEqual(contig.Sequences[i].Position, deserializedContig.Sequences[i].Position);
}
}
catch (Exception)
{
Assert.Fail();
}
finally
{
if (stream != null)
{
stream.Close();
stream = null;
}
}
}
/// <summary>
/// Aligns reads to contigs using kmer method of alignment.
/// </summary>
/// <param name="contigs">List of contig sequences.</param>
/// <param name="reads">List of read sequences.</param>
/// <param name="kmerLength">Kmer Length.</param>
/// <returns>List of Contig.</returns>
public static IList<Contig> ReadContigAlignment(IList<ISequence> contigs, IList<ISequence> reads, int kmerLength)
{
KmerIndexerDictionary map = SequenceToKmerBuilder.BuildKmerDictionary(reads, kmerLength);
IList<ContigIndex> contigDatas;
contigDatas = contigs.AsParallel().Select(contig =>
{
IEnumerable<ISequence> kmers = SequenceToKmerBuilder.GetKmerSequences(contig, kmerLength);
ContigIndex index = new ContigIndex(contig);
foreach (ISequence kmer in kmers)
{
IList<KmerIndexer> positions;
if (map.TryGetValue(kmer, out positions) ||
map.TryGetValue(kmer.GetReverseComplementedSequence(), out positions))
{
index.ContigReadMatchIndexes.Add(positions);
}
else
{
index.ContigReadMatchIndexes.Add(new List<KmerIndexer>());
}
}
return index;
}).ToList();
return contigDatas.Select(contigData =>
{
IList<Task<IList<ReadMap>>> tasks =
new List<Task<IList<ReadMap>>>();
// Stores information about contigs for which tasks has been generated.
IList<long> visitedReads = new List<long>();
// Creates Task for every read in nodes for a given contig.
for (int index = 0; index < contigData.ContigReadMatchIndexes.Count; index++)
{
int readPosition = index;
foreach (KmerIndexer kmer in contigData.ContigReadMatchIndexes[index])
{
long contigIndex = kmer.SequenceIndex;
if (!visitedReads.Contains(contigIndex))
{
visitedReads.Add(contigIndex);
tasks.Add(
Task<IList<ReadMap>>.Factory.StartNew(t => MapRead(readPosition, contigData.ContigReadMatchIndexes, contigIndex, kmerLength), TaskCreationOptions.AttachedToParent));
}
}
}
Contig contigOutputStructure = new Contig();
contigOutputStructure.Consensus = contigData.ContigSequence;
for (int index = 0; index < visitedReads.Count; index++)
{
foreach (ReadMap maps in tasks[index].Result)
{
Contig.AssembledSequence assembledSeq = new Contig.AssembledSequence()
{
Length = maps.Length,
Position = maps.StartPositionOfContig,
ReadPosition = maps.StartPositionOfRead,
Sequence = reads.ElementAt(visitedReads[index])
};
if (new string(
contigOutputStructure.Consensus.GetSubSequence(
assembledSeq.Position, assembledSeq.Length).Select(a => (char)a).ToArray()).
Equals(new string(assembledSeq.Sequence.GetSubSequence(assembledSeq.ReadPosition, assembledSeq.Length)
.Select(a => (char)a).ToArray())))
{
assembledSeq.IsComplemented = false;
assembledSeq.IsReversed = false;
}
else
{
assembledSeq.IsComplemented = true;
assembledSeq.IsReversed = true;
}
contigOutputStructure.Sequences.Add(assembledSeq);
}
}
return contigOutputStructure;
}).ToList();
}
