public CCSDataSet(Sequence ccs, List<Sequence> subreads, BWAPairwiseAlignment aln)
{
// Call Variants
var results = VariantCaller.LeftAlignIndelsAndCallVariants (aln);
Variants = results.Item2;
Variants.ForEach (p => {
p.StartPosition += aln.AlignedSAMSequence.Pos;
p.RefName = aln.Reference;
});
// Get the ungapped reference
var ccs_aligned = results.Item1.SecondSequence;
// Generate an ungapped reference
var refseq = results.Item1.PairwiseAlignedSequences.First().FirstSequence;
var reference = new Sequence (DnaAlphabet.Instance,
refseq.Where (bp => bp != '-').ToArray (),
false)
{ ID = aln.Reference + "/" + aln.AlignedSAMSequence.Pos + "-" + aln.AlignedSAMSequence.RefEndPos };
var reference_rc = reference.GetReverseComplementedSequence () as Sequence;
Reads = subreads.Select(z => new AlignedSequence(z, reference, reference_rc)).Where( x => (Math.Abs((x.Sequence.Length - reference.Count)) / (double) reference.Count) < 0.2).ToList();
Reads.Insert (0, new AlignedSequence (reference, reference, reference_rc));
ccs.ID = "CCS";
Reads.Insert(1, new AlignedSequence(ccs, reference, reference_rc));
// Now to rejigger the alignments
int[] metaOffSets = Enumerable.Range(0, (int)reference.Count).
Select (i => Reads.Max( r => r.GetInsertsBeforePosition(i))).
ToArray();
Reads.ForEach(r => {
r.ReJiggerAlignment(metaOffSets);
Console.WriteLine(r.Sequence);});
CCS = ccs;
Reference = reference;
Aln = aln;
}
/// <summary>
/// Calculation method.
/// </summary>
/// <param name="chains">
/// The chains.
/// </param>
/// <param name="calculators">
/// The calculators.
/// </param>
/// <param name="links">
/// The links.
/// </param>
/// <param name="rotate">
/// The rotate flag.
/// </param>
/// <param name="complementary">
/// The complementary flag.
/// </param>
/// <param name="rotationLength">
/// The rotation length.
/// </param>
/// <returns>
/// The <see cref="T:double[][]"/>.
/// </returns>
public static double[][] Calculate(Chain[][] chains, IFullCalculator[] calculators, Link[] links, bool rotate, bool complementary, uint? rotationLength)
{
var characteristics = new double[chains.Length][];
for (int i = 0; i < chains.Length; i++)
{
characteristics[i] = new double[calculators.Length];
for (int j = 0; j < calculators.Length; j++)
{
if (complementary)
{
var sourceSequence = new Sequence(Alphabets.DNA, chains[i][j].ToString());
var complementarySequence = sourceSequence.GetReverseComplementedSequence();
chains[i][j] = new Chain(complementarySequence.ConvertToString());
}
if (rotate)
{
var building = chains[i][j].Building.Rotate(rotationLength ?? 0);
var newSequence = building.Select(t => new ValueInt(t)).Cast<IBaseObject>().ToList();
chains[i][j] = new Chain(newSequence);
}
chains[i][j].FillIntervalManagers();
characteristics[i][j] = calculators[j].Calculate(chains[i][j], links[j]);
}
}
return characteristics;
}
public static void TestTrickyQVInversions() {
// This will be hard because normally flip the QV value for a homopolymer, but in this case we won't.
// Note the whole notion of flipping is poorly defined.
string seq1seq = "ATTGC";
string seq2seq = "ATAGC";
int[] seq2qual = new int[] { 30, 30, 2, 30, 30 };
var refseq = new Sequence(DnaAlphabet.Instance, seq1seq);
var query = new Sequence (DnaAlphabet.Instance, seq2seq);
var s1rc = refseq.GetReverseComplementedSequence ();
var s2rc = query.GetReverseComplementedSequence ();
NeedlemanWunschAligner aligner = new NeedlemanWunschAligner ();
var aln = aligner.Align (s1rc, s2rc).First();
VariantCallTests.ConvertAlignedSequenceToQualSeq (aln, seq2qual.Reverse ().ToArray ());
aln.PairwiseAlignedSequences [0].Sequences [1].MarkAsReverseComplement ();
var variants = VariantCaller.CallVariants (aln);
Assert.AreEqual (1, variants.Count);
var variant = variants.First ();
Assert.AreEqual (VariantType.SNP, variant.Type);
Assert.AreEqual (2, variant.QV);
var vs = variant as SNPVariant;
Assert.AreEqual ('T', vs.AltBP);
Assert.AreEqual ('A', vs.RefBP);
}
public static void TestReverseComplement1BPIndelCall() {
string seq1seq = "ATACCCCTTGCGC";
string seq2seq = "ATA-CCCTTGCGC".Replace("-", String.Empty);
int[] seq2qual = new int[] { 30, 30, 30, 2, 30, 30, 30, 30, 30, 30, 30, 30 };
var refseq = new Sequence(DnaAlphabet.Instance, seq1seq);
var query = new Sequence (DnaAlphabet.Instance, seq2seq);
var s1rc = refseq.GetReverseComplementedSequence ();
var s2rc = query.GetReverseComplementedSequence ();
NeedlemanWunschAligner aligner = new NeedlemanWunschAligner ();
var aln = aligner.Align (s1rc, s2rc).First();
VariantCallTests.ConvertAlignedSequenceToQualSeq (aln, seq2qual.Reverse ().ToArray ());
aln.PairwiseAlignedSequences [0].Sequences [1].MarkAsReverseComplement ();
var variants = VariantCaller.CallVariants (aln);
Assert.AreEqual (variants.Count, 1);
var variant = variants.First ();
Assert.AreEqual (2, variant.QV);
Assert.AreEqual (5, variant.StartPosition);
Assert.AreEqual (VariantType.INDEL, variant.Type);
var vi = variant as IndelVariant;
Assert.AreEqual (IndelType.Deletion, vi.InsertionOrDeletion);
Assert.AreEqual ('G', vi.HomopolymerBase);
Assert.AreEqual (1, vi.Length);
Assert.AreEqual (4, vi.HomopolymerLengthInReference);
Assert.AreEqual (true, vi.InHomopolymer);
Assert.AreEqual ("G", vi.InsertedOrDeletedBases);
Assert.AreEqual (false, vi.AtEndOfAlignment);
Assert.AreEqual (6, vi.EndPosition);
}
