public static void Test1BPDeletionCall()
{
string seq1seq = "ATACCCCTT";
string seq2seq = "ATA-CCCTT".Replace("-", String.Empty);
int[] seq2qual = new int[] { 30, 30, 30, 2, 30, 30, 30, 30 };
var refseq = new Sequence(AmbiguousDnaAlphabet.Instance, seq1seq, false);
var query = new Sequence(AmbiguousDnaAlphabet.Instance, seq2seq, false);
NeedlemanWunschAligner aligner = new NeedlemanWunschAligner();
var aln = aligner.Align(refseq, query).First();
// Need to add in the QV Values.
ConvertAlignedSequenceToQualSeq(aln, seq2qual);
var variants = VariantCaller.CallVariants(aln);
Assert.AreEqual(variants.Count, 1);
var variant = variants.First();
Assert.AreEqual(2, variant.QV);
Assert.AreEqual(2, variant.StartPosition);
Assert.AreEqual(VariantType.INDEL, variant.Type);
var vi = variant as IndelVariant;
Assert.AreEqual("C", vi.InsertedOrDeletedBases);
Assert.AreEqual('C', vi.HomopolymerBase);
Assert.AreEqual(4, vi.HomopolymerLengthInReference);
Assert.AreEqual(true, vi.InHomopolymer);
Assert.AreEqual(vi.InsertionOrDeletion, IndelType.Deletion);
}
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 TestSNPCallAtStart()
{
string seq1seq = "CTCCCCCTT";
string seq2seq = "TTCCCCCTT";
int[] seq2qual = new int[] { 10, 30, 30, 30, 5, 3, 30, 30, 10 };
var refseq = new Sequence(DnaAlphabet.Instance, seq1seq);
var query = new Sequence(DnaAlphabet.Instance, seq2seq);
NeedlemanWunschAligner aligner = new NeedlemanWunschAligner();
var aln = aligner.Align(refseq, query).First();
ConvertAlignedSequenceToQualSeq(aln, seq2qual);
var variants = VariantCaller.CallVariants(aln);
Assert.AreEqual(variants.Count, 1);
var variant = variants.First();
Assert.AreEqual(10, variant.QV);
Assert.AreEqual(0, variant.StartPosition);
Assert.AreEqual(variant.Type, VariantType.SNP);
var vi = variant as SNPVariant;
Assert.AreEqual(1, vi.Length);
Assert.AreEqual('T', vi.AltBP);
Assert.AreEqual('C', vi.RefBP);
Assert.AreEqual(VariantType.SNP, vi.Type);
Assert.AreEqual(true, vi.AtEndOfAlignment);
}
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);
}
public static void Main(string[] args)
{
try {
PlatformManager.Services.MaxSequenceSize = int.MaxValue;
PlatformManager.Services.DefaultBufferSize = 4096;
PlatformManager.Services.Is64BitProcessType = true;
if (args.Length > 3)
{
Console.WriteLine("Too many arguments");
DisplayHelp();
}
else if (args.Length < 2)
{
Console.WriteLine("Not enough arguments");
DisplayHelp();
}
else if (args [0] == "h" || args [0] == "help" || args [0] == "?" || args [0] == "-h")
{
DisplayHelp();
}
else
{
string bam_name = args [0];
string out_dir = args [1];
string ref_name = args.Length > 2 ? args [2] : null;
if (!File.Exists(bam_name))
{
Console.WriteLine("Can't find file: " + bam_name);
return;
}
if (ref_name != null && !File.Exists(ref_name))
{
Console.WriteLine("Can't find file: " + ref_name);
return;
}
if (Directory.Exists(out_dir))
{
Console.WriteLine("The output directory already exists, please specify a new directory or delete the old one.");
return;
}
Directory.CreateDirectory(out_dir);
List <CCSReadMetricsOutputter> outputters = new List <CCSReadMetricsOutputter> ()
{
new ZmwOutputFile(out_dir),
new ZScoreOutputter(out_dir),
new VariantOutputter(out_dir),
new SNROutputFile(out_dir),
new QVCalibration(out_dir)
};
ISequenceParser reader;
if (bam_name.EndsWith(".fastq", StringComparison.OrdinalIgnoreCase))
{
reader = new FastQCCSReader();
}
else
{
reader = new PacBioCCSBamReader();
}
BWAPairwiseAligner bwa = null;
bool callVariants = ref_name != null;
if (callVariants)
{
bwa = new BWAPairwiseAligner(ref_name, false);
}
// Produce aligned reads with variants called in parallel.
var reads = new BlockingCollection <Tuple <PacBioCCSRead, BWAPairwiseAlignment, List <Variant> > >();
Task producer = Task.Factory.StartNew(() =>
{
try
{
Parallel.ForEach(reader.Parse(bam_name), y => {
var z = y as PacBioCCSRead;
try {
BWAPairwiseAlignment aln = null;
List <Variant> variants = null;
if (callVariants)
{
aln = bwa.AlignRead(z.Sequence) as BWAPairwiseAlignment;
if (aln != null)
{
variants = VariantCaller.CallVariants(aln);
variants.ForEach(p => {
p.StartPosition += aln.AlignedSAMSequence.Pos;
p.RefName = aln.Reference;
});
}
}
var res = new Tuple <PacBioCCSRead, BWAPairwiseAlignment, List <Variant> >(z, aln, variants);
reads.Add(res);
}
catch (Exception thrown) {
Console.WriteLine("CCS READ FAIL: " + z.Sequence.ID);
Console.WriteLine(thrown.Message);
}
});
} catch (Exception thrown) {
Console.WriteLine("Could not parse BAM file: " + thrown.Message);
while (thrown.InnerException != null)
{
Console.WriteLine(thrown.InnerException.Message);
thrown = thrown.InnerException;
}
}
reads.CompleteAdding();
});
// Consume them into output files.
foreach (var r in reads.GetConsumingEnumerable())
{
foreach (var outputter in outputters)
{
outputter.ConsumeCCSRead(r.Item1, r.Item2, r.Item3);
}
}
// throw any exceptions (this should be used after putting the consumer on a separate thread)
producer.Wait();
// Close the files
outputters.ForEach(z => z.Finish());
}
}
catch (DllNotFoundException thrown) {
Console.WriteLine("Error thrown when attempting to generate the CCS results.");
Console.WriteLine("A shared library was not found. To solve this, please add the folder" +
" with the downloaded files libbwasharp and libMonoPosixHelper" +
"to your environmental variables (LD_LIBRARY_PATH on Ubuntu, DYLD_LIBRARY_PATH on Mac OS X).");
Console.WriteLine("Error: " + thrown.Message);
Console.WriteLine(thrown.StackTrace);
}
catch (Exception thrown) {
Console.WriteLine("Error thrown when attempting to generate the CCS results");
Console.WriteLine("Error: " + thrown.Message);
Console.WriteLine(thrown.StackTrace);
while (thrown.InnerException != null)
{
Console.WriteLine("Inner Exception: " + thrown.InnerException.Message);
thrown = thrown.InnerException;
}
}
}
