public void Initialize()
{
VcfWriterConfig config = new VcfWriterConfig
{
DepthFilterThreshold = 500,
VariantQualityFilterThreshold = 20,
StrandBiasFilterThreshold = 0.5f,
FrequencyFilterThreshold = 0.007f,
MinFrequencyThreshold = 0.007f,
ShouldOutputNoCallFraction = true,
ShouldOutputStrandBiasAndNoiseLevel = true,
ShouldFilterOnlyOneStrandCoverage = true,
EstimatedBaseCallQuality = _estimatedBaseCallQuality,
//AllowMultipleVcfLinesPerLoci = true
};
_formatter = new VcfFormatter(config);
_v1 = TestHelper.CreatePassingVariant(false);
_v2 = TestHelper.CreatePassingVariant(false);
_v3 = TestHelper.CreatePassingVariant(false);
}
public void FilterHeader()
{
var outputFilePath = Path.Combine(UnitTestPaths.TestDataDirectory, "VcfFileWriterTests_SDS-18.vcf");
File.Delete(outputFilePath);
var context = new VcfWriterInputContext
{
CommandLine = "myCommandLine",
SampleName = "mySample",
ReferenceName = "myReference",
ContigsByChr = new List <Tuple <string, long> >
{
new Tuple <string, long>("chr1", 10001),
new Tuple <string, long>("chrX", 500)
}
};
// Variant strand bias too high or coverage on only one strand
var config = new VcfWriterConfig
{
DepthFilterThreshold = 500,
QscoreFilterThreshold = 20,
StrandBiasFilterThreshold = 0.5f,
FrequencyFilterThreshold = 0.007f,
ShouldOutputNoCallFraction = true,
ShouldOutputStrandBiasAndNoiseLevel = true,
ShouldFilterOnlyOneStrandCoverage = true,
EstimatedBaseCallQuality = 23
};
var writer = new VcfFileWriter(outputFilePath, config, context);
writer.WriteHeader();
writer.Write(_defaultCandidates);
writer.Dispose();
VcfHeaderFormatTester(config, outputFilePath);
}
public void WriteANbhd()
{
var outputFilePath = Path.Combine(TestPaths.LocalTestDataDirectory, "PhasedVcfFileNbhdWriterTest.vcf");
var inputFilePath = Path.Combine(TestPaths.LocalTestDataDirectory, "MergerInput.vcf");
var expectedFilePath = Path.Combine(TestPaths.LocalTestDataDirectory, "MergerOutput.vcf");
File.Delete(outputFilePath);
var context = new VcfWriterInputContext
{
QuotedCommandLineString = "myCommandLine",
SampleName = "mySample",
ReferenceName = "myReference",
ContigsByChr = new List <Tuple <string, long> >
{
new Tuple <string, long>("chr1", 10001),
new Tuple <string, long>("chrX", 500)
}
};
var config = new VcfWriterConfig
{
DepthFilterThreshold = 500,
VariantQualityFilterThreshold = 30,
FrequencyFilterThreshold = 0.007f,
ShouldOutputNoCallFraction = true,
ShouldOutputStrandBiasAndNoiseLevel = true,
EstimatedBaseCallQuality = 23,
PloidyModel = PloidyModel.Somatic,
AllowMultipleVcfLinesPerLoci = true
};
var writer = new PhasedVcfWriter(outputFilePath, config, new VcfWriterInputContext(), new List <string>()
{
}, null);
var reader = new VcfReader(inputFilePath, true);
//set up the original variants
var originalVcfVariant1 = TestHelper.CreateDummyAllele("chr2", 116380048, "A", "New", 1000, 156);
var originalVcfVariant2 = TestHelper.CreateDummyAllele("chr2", 116380048, "AAA", "New", 1000, 156);
var originalVcfVariant4 = TestHelper.CreateDummyAllele("chr7", 116380051, "A", "New", 1000, 156);
var originalVcfVariant5 = TestHelper.CreateDummyAllele("chr7", 116380052, "AC", "New", 1000, 156);
var vs1 = new VariantSite((originalVcfVariant1));
var vs2 = new VariantSite((originalVcfVariant2));
var vs4 = new VariantSite((originalVcfVariant4));
var vs5 = new VariantSite((originalVcfVariant5));
//have to replace variants at positon 116380048 (we call two new MNVS here)
var nbhd1 = new VcfNeighborhood(new VariantCallingParameters(), 0, "chr2", vs1, vs2, "");
nbhd1.SetRangeOfInterest();
//have to replace variants at positon 116380051 and 52 (we call one new MNV at 51)
var nbhd2 = new VcfNeighborhood(new VariantCallingParameters(), 0, "chr7", vs4, vs5, "");
nbhd2.SetRangeOfInterest();
VcfMerger merger = new VcfMerger(reader);
List <CalledAllele> allelesPastNbh = new List <CalledAllele>();
nbhd1.CalledVariants = new Dictionary <int, List <CalledAllele> > {
{ originalVcfVariant1.ReferencePosition, new List <CalledAllele> {
originalVcfVariant1, originalVcfVariant2
} }
};
nbhd2.CalledVariants = new Dictionary <int, List <CalledAllele> > {
{ originalVcfVariant4.ReferencePosition, new List <CalledAllele> {
originalVcfVariant4
} }
};
allelesPastNbh = merger.WriteVariantsUptoChr(writer, allelesPastNbh, nbhd1.ReferenceName);
allelesPastNbh = merger.WriteVariantsUptoIncludingNbhd(nbhd1, writer, allelesPastNbh);
allelesPastNbh = merger.WriteVariantsUptoChr(writer, allelesPastNbh, nbhd2.ReferenceName);
allelesPastNbh = merger.WriteVariantsUptoIncludingNbhd(nbhd2, writer, allelesPastNbh);
merger.WriteRemainingVariants(writer, allelesPastNbh);
writer.Dispose();
var expectedLines = File.ReadLines(expectedFilePath).ToList();
var outputLines = File.ReadLines(outputFilePath).ToList();
Assert.Equal(expectedLines.Count(), outputLines.Count());
for (int i = 0; i < expectedLines.Count; i++)
{
Assert.Equal(expectedLines[i], outputLines[i]);
}
}
private void VcfHeaderFormatTester(VcfWriterConfig config, string outputFile)
{
// Time to read the header
var testFile = File.ReadAllLines(outputFile);
bool formatLowDP = false, formatQ = false, formatSB = false;
foreach (var x in testFile.Where(x => Regex.IsMatch(x, "##FILTER=")))
{
switch (x.Split(',')[0])
{
case "##FILTER=<ID=LowDP":
Assert.True(Regex.IsMatch(x, "^##FILTER=<ID=LowDP,Description=\"Low coverage \\(DP tag\\), therefore no genotype called\">$"));
formatLowDP = true;
break;
case "##FILTER=<ID=SB":
if (config.StrandBiasFilterThreshold.HasValue && config.ShouldFilterOnlyOneStrandCoverage)
{
Assert.True(Regex.IsMatch(x, "^##FILTER=<ID=SB,Description=\"(Variant strand bias too high or coverage on only one strand)\">$"));
}
else if (config.StrandBiasFilterThreshold.HasValue)
{
Assert.True(Regex.IsMatch(x, "^##FILTER=<ID=SB,Description=\"(Variant strand bias too high)\">$"));
}
else if (config.ShouldFilterOnlyOneStrandCoverage)
{
Assert.True(Regex.IsMatch(x, "^##FILTER=<ID=SB,Description=\"(Variant support on only one strand)\">$"));
}
else
{
Assert.True(false, "StrandBias filter header does not match any expected filter.");
}
formatSB = true;
break;
default:
if (Regex.IsMatch(x, string.Format("##FILTER=<ID=q{0}", config.QscoreFilterThreshold)))
{
Assert.True(Regex.IsMatch(x, string.Format("^##FILTER=<ID=q{0},Description=\"Quality below {0}\">$", config.QscoreFilterThreshold)));
formatQ = true;
}
else
{
Assert.True(false, "A filter is listed which does not match any of the specified filters.");
}
break;
}
}
if (config.QscoreFilterThreshold > 0)
{
Assert.True(formatQ);
}
if (config.DepthFilterThreshold > 0)
{
Assert.True(formatLowDP);
}
if (config.ShouldOutputStrandBiasAndNoiseLevel)
{
Assert.True(formatSB);
}
}
public void DataFormatCheck()
{
var outputFilePath = Path.Combine(UnitTestPaths.TestDataDirectory, "VcfFileWriterTests_SDS-23.vcf");
File.Delete(outputFilePath);
var context = new VcfWriterInputContext
{
CommandLine = "myCommandLine",
SampleName = "mySample",
ReferenceName = "myReference",
ContigsByChr = new List <Tuple <string, long> >
{
new Tuple <string, long>("chr1", 10001),
new Tuple <string, long>("chrX", 500)
}
};
var config = new VcfWriterConfig
{
DepthFilterThreshold = 500,
QscoreFilterThreshold = 20,
StrandBiasFilterThreshold = 0.5f,
FrequencyFilterThreshold = 0.007f,
ShouldOutputNoCallFraction = true,
ShouldOutputStrandBiasAndNoiseLevel = true,
ShouldFilterOnlyOneStrandCoverage = true,
EstimatedBaseCallQuality = 23
};
var writer = new VcfFileWriter(outputFilePath, config, context);
writer.WriteHeader();
writer.Write(_defaultCandidates);
writer.Dispose();
var testFile = File.ReadAllLines(outputFilePath);
var formatList = string.Empty;
bool caseNL = false, caseSB = false, caseNC = false;
foreach (var x in testFile)
{
if (Regex.IsMatch(x, "^##FORMAT"))
{
var formatField = x.Split(',')[0].Substring(13);
switch (formatField)
{
case "NL":
if (config.ShouldOutputStrandBiasAndNoiseLevel)
{
caseNL = true;
}
break;
case "SB":
if (config.ShouldOutputStrandBiasAndNoiseLevel)
{
caseSB = true;
}
break;
case "NC":
if (config.ShouldOutputNoCallFraction)
{
caseNC = true;
}
break;
}
if (formatList == string.Empty)
{
formatList = x.Split(',')[0].Substring(13);
}
else
{
formatList += ":" + x.Split(',')[0].Substring(13);
}
}
if (Regex.IsMatch(x, "^chr\\d+\t"))
{
var y = x.Split('\t');
Assert.True(Regex.IsMatch(y[8], formatList));
}
}
if ((!config.ShouldOutputStrandBiasAndNoiseLevel && caseNL) ||
(config.ShouldOutputStrandBiasAndNoiseLevel && !caseNL))
{
Assert.True(false, "Incorrect setting for ShouldOutputStrandBiasAndNoiseLevel and NL format");
}
if ((!config.ShouldOutputStrandBiasAndNoiseLevel && caseSB) ||
(config.ShouldOutputStrandBiasAndNoiseLevel && !caseSB))
{
Assert.True(false, "Incorrect setting for ShouldOutputStrandBiasAndNoiseLevel and SB format");
}
if ((!config.ShouldOutputNoCallFraction && caseNC) || (config.ShouldOutputNoCallFraction && !caseNC))
{
Assert.True(false, "Incorrect setting for NoCall and NC format");
}
}
public void InfoFormatHeader()
{
var outputFilePath = Path.Combine(UnitTestPaths.TestDataDirectory, "VcfFileWriterTests_SDS-17.vcf");
File.Delete(outputFilePath);
var context = new VcfWriterInputContext
{
CommandLine = "myCommandLine",
SampleName = "mySample",
ReferenceName = "myReference",
ContigsByChr = new List <Tuple <string, long> >
{
new Tuple <string, long>("chr1", 10001),
new Tuple <string, long>("chrX", 500)
}
};
var config = new VcfWriterConfig
{
DepthFilterThreshold = 500,
QscoreFilterThreshold = 20,
StrandBiasFilterThreshold = 0.5f,
FrequencyFilterThreshold = 0.007f,
ShouldOutputNoCallFraction = true,
ShouldOutputStrandBiasAndNoiseLevel = true,
ShouldFilterOnlyOneStrandCoverage = true,
EstimatedBaseCallQuality = 23
};
var writer = new VcfFileWriter(outputFilePath, config, context);
writer.WriteHeader();
writer.Write(_defaultCandidates);
writer.Dispose();
// Time to read the header
var testFile = File.ReadAllLines(outputFilePath);
bool formatNL = false, formatSB = false, formatNC = false;
foreach (var x in testFile)
{
if (Regex.IsMatch(x, "##INFO="))
{
switch (x.Split(',')[0])
{
case "##INFO=<ID=DP":
Assert.True(Regex.IsMatch(x, "^##INFO=<ID=DP,Number=1,Type=Integer,Description=\"Total Depth\">$"));
break;
case "##INFO=<ID=TI":
Assert.True(Regex.IsMatch(x, "^##INFO=<ID=TI,Number=\\.,Type=String,Description=\"Transcript ID\">$"));
break;
case "##INFO=<ID=GI":
Assert.True(Regex.IsMatch(x, "^##INFO=<ID=GI,Number=\\.,Type=String,Description=\"Gene ID\">$"));
break;
case "##INFO=<ID=EXON":
Assert.True(Regex.IsMatch(x, "^##INFO=<ID=EXON,Number=0,Type=Flag,Description=\"Exon Region\">$"));
break;
case "##INFO=<ID=FC":
Assert.True(Regex.IsMatch(x, "^##INFO=<ID=FC,Number=\\.,Type=String,Description=\"Functional Consequence\">$"));
break;
default:
Assert.True(false, "An info is listed which does not match any from the req.`");
break;
}
}
else if (Regex.IsMatch(x, "##FORMAT="))
{
switch (x.Split(',')[0])
{
case "##FORMAT=<ID=GT":
Assert.True(Regex.IsMatch(x, "^##FORMAT=<ID=GT,Number=1,Type=String,Description=\"Genotype\">$"));
break;
case "##FORMAT=<ID=GQ":
Assert.True(Regex.IsMatch(x, "^##FORMAT=<ID=GQ,Number=1,Type=Integer,Description=\"Genotype Quality\">$"));
break;
case "##FORMAT=<ID=AD":
Assert.True(Regex.IsMatch(x, "^##FORMAT=<ID=AD,Number=\\.,Type=Integer,Description=\"Allele Depth\">$"));
break;
case "##FORMAT=<ID=VF":
Assert.True(Regex.IsMatch(x, "^##FORMAT=<ID=VF,Number=1,Type=Float,Description=\"Variant Frequency\">$"));
break;
case "##FORMAT=<ID=NL":
Assert.True(Regex.IsMatch(x, "^##FORMAT=<ID=NL,Number=1,Type=Integer,Description=\"Applied BaseCall Noise Level\">$"));
formatNL = true;
break;
case "##FORMAT=<ID=SB":
Assert.True(Regex.IsMatch(x, "^##FORMAT=<ID=SB,Number=1,Type=Float,Description=\"StrandBias Score\">$"));
formatSB = true;
break;
case "##FORMAT=<ID=NC":
Assert.True(Regex.IsMatch(x, "^##FORMAT=<ID=NC,Number=1,Type=Float,Description=\"Fraction of bases which were uncalled or with basecall quality below the minimum threshold\">$"));
formatNC = true;
break;
default:
Assert.True(false, "A format is listed which does not match any of those listed for the req.");
break;
}
}
}
if (config.ShouldOutputStrandBiasAndNoiseLevel)
{
Assert.True(formatNL);
}
if (config.ShouldOutputStrandBiasAndNoiseLevel)
{
Assert.True(formatSB);
}
if (config.ShouldOutputNoCallFraction)
{
Assert.True(formatNC);
}
}
public void FilterHeader()
{
var outputFilePath = Path.Combine(TestPaths.LocalTestDataDirectory, "PhasedVcfFileWriterTests.vcf");
File.Delete(outputFilePath);
var context = new VcfWriterInputContext
{
QuotedCommandLineString = "myCommandLine",
SampleName = "mySample",
ReferenceName = "myReference",
ContigsByChr = new List <Tuple <string, long> >
{
new Tuple <string, long>("chr1", 10001),
new Tuple <string, long>("chrX", 500)
}
};
// Variant strand bias too high or coverage on only one strand
var config = new VcfWriterConfig
{
DepthFilterThreshold = 500,
VariantQualityFilterThreshold = 30,
FrequencyFilterThreshold = 0.007f,
ShouldOutputNoCallFraction = true,
ShouldOutputStrandBiasAndNoiseLevel = true,
EstimatedBaseCallQuality = 23,
PloidyModel = PloidyModel.Diploid,
};
//note, scylla has no SB or RMxN or R8 filters.
var variants = new List <CalledAllele>
{
TestHelper.CreateDummyAllele("chrX", 123, "A", "C", 1000, 156),
TestHelper.CreateDummyAllele("chr10", 124, "A", "C", 1000, 156),
};
var originalHeader = new List <string>
{
"##fileformat=VCFv4.1",
"##fileDate=20160620",
"##source=Pisces 1.0.0.0",
"##Pisces_cmdline=\"-B KRAS_42_S1.bam -g -MinimumFrequency 0.01 -MinBaseCallQuality 21 -MaxVariantQScore 100 -MinCoverage 300 -MaxAcceptableStrandBiasFilter 0.5 -MinVariantQScore 20 -VariantQualityFilter 20 -gVCF true -CallMNVs True -out \\myout",
"##reference=WholeGenomeFASTA",
"##INFO=<ID=DP,Number=1,Type=Integer,Description=\"Total Depth\">",
"##FILTER=<ID=q20,Description=\"Quality score less than 20\">",
"##FILTER=<ID=SB,Description=\"Variant strand bias too high\">",
"##FILTER=<ID=R5x9,Description=\"Repeats of part or all of the variant allele (max repeat length 5) in the reference greater than or equal to 9\">",
"##FORMAT=<ID=GT,Number=1,Type=String,Description=\"Genotype\">",
"##FORMAT=<ID=GQ,Number=1,Type=Integer,Description=\"Genotype Quality\">",
"#CHROM POS ID REF ALT QUAL FILTER INFO FORMAT HD700n560_miseq1_S7.bam"
};
var writer = new PhasedVcfWriter(outputFilePath, config, new VcfWriterInputContext(), originalHeader, null);
writer.WriteHeader();
writer.Write(variants);
writer.Dispose();
VcfReader reader = new VcfReader(outputFilePath);
List <string> writtenHeader = reader.HeaderLines;
reader.Dispose();
var expectedHeader1 = new List <string>
{
"##fileformat=VCFv4.1",
"##fileDate=20160620",
"##source=Pisces 1.0.0.0",
"##Pisces_cmdline=\"-B KRAS_42_S1.bam -g -MinimumFrequency 0.01 -MinBaseCallQuality 21 -MaxVariantQScore 100 -MinCoverage 300 -MaxAcceptableStrandBiasFilter 0.5 -MinVariantQScore 20 -VariantQualityFilter 20 -gVCF true -CallMNVs True -out \\myout",
"##VariantPhaser=Scylla 1.0.0.0",
"##reference=WholeGenomeFASTA",
"##INFO=<ID=DP,Number=1,Type=Integer,Description=\"Total Depth\">",
"##FILTER=<ID=q20,Description=\"Quality score less than 20\">",
"##FILTER=<ID=SB,Description=\"Variant strand bias too high\">",
"##FILTER=<ID=R5x9,Description=\"Repeats of part or all of the variant allele (max repeat length 5) in the reference greater than or equal to 9\">",
"##FILTER=<ID=q30,Description=\"Quality score less than 30, by Scylla\">",
"##FILTER=<ID=LowDP,Description=\"Low coverage (DP tag), therefore no genotype called, by Scylla\">",
"##FILTER=<ID=LowVariantFreq,Description=\"Variant frequency less than 0.0070, by Scylla\">",
"##FILTER=<ID=MultiAllelicSite,Description=\"Variant does not conform to diploid model, by Scylla\">",
"##FORMAT=<ID=GT,Number=1,Type=String,Description=\"Genotype\">",
"##FORMAT=<ID=GQ,Number=1,Type=Integer,Description=\"Genotype Quality\">",
"#CHROM POS ID REF ALT QUAL FILTER INFO FORMAT HD700n560_miseq1_S7.bam"
};
Assert.Equal(expectedHeader1.Count, writtenHeader.Count);
for (int i = 0; i < expectedHeader1.Count; i++)
{
//let version numbers differ
if (expectedHeader1[i].StartsWith("##VariantPhaser=Scylla"))
{
Assert.True(writtenHeader[i].StartsWith("##VariantPhaser=Scylla"));
continue;
}
Assert.Equal(expectedHeader1[i], writtenHeader[i]);
}
config = new VcfWriterConfig
{
DepthFilterThreshold = 500,
VariantQualityFilterThreshold = 22,
FrequencyFilterThreshold = 0.007f,
EstimatedBaseCallQuality = 23,
PloidyModel = PloidyModel.Somatic,
};
originalHeader = new List <string>
{
"##fileformat=VCFv4.1",
"##fileDate=20160620",
"##source=Pisces 1.0.0.0",
"##Pisces_cmdline=\"-B KRAS_42_S1.bam -g -MinimumFrequency 0.01 -MinBaseCallQuality 21 -MaxVariantQScore 100 -MinCoverage 300 -MaxAcceptableStrandBiasFilter 0.5 -MinVariantQScore 20 -VariantQualityFilter 20 -gVCF true -CallMNVs True -out \\myout",
"##reference=WholeGenomeFASTA",
"##INFO=<ID=DP,Number=1,Type=Integer,Description=\"Total Depth\">",
"##FORMAT=<ID=GT,Number=1,Type=String,Description=\"Genotype\">",
"##FORMAT=<ID=GQ,Number=1,Type=Integer,Description=\"Genotype Quality\">",
"#CHROM POS ID REF ALT QUAL FILTER INFO FORMAT HD700n560_miseq1_S7.bam"
};
writer = new PhasedVcfWriter(outputFilePath, config, new VcfWriterInputContext(), originalHeader, null);
var expectedHeader2 = new List <string>
{
"##fileformat=VCFv4.1",
"##fileDate=20160620",
"##source=Pisces 1.0.0.0",
"##Pisces_cmdline=\"-B KRAS_42_S1.bam -g -MinimumFrequency 0.01 -MinBaseCallQuality 21 -MaxVariantQScore 100 -MinCoverage 300 -MaxAcceptableStrandBiasFilter 0.5 -MinVariantQScore 20 -VariantQualityFilter 20 -gVCF true -CallMNVs True -out \\myout",
"##VariantPhaser=Scylla 1.0.0.0",
"##reference=WholeGenomeFASTA",
"##INFO=<ID=DP,Number=1,Type=Integer,Description=\"Total Depth\">",
"##FORMAT=<ID=GT,Number=1,Type=String,Description=\"Genotype\">",
"##FORMAT=<ID=GQ,Number=1,Type=Integer,Description=\"Genotype Quality\">",
"##FILTER=<ID=q22,Description=\"Quality score less than 22, by Scylla\">",
"##FILTER=<ID=LowDP,Description=\"Low coverage (DP tag), therefore no genotype called, by Scylla\">",
"##FILTER=<ID=LowVariantFreq,Description=\"Variant frequency less than 0.0070, by Scylla\">",
"#CHROM POS ID REF ALT QUAL FILTER INFO FORMAT HD700n560_miseq1_S7.bam"
};
writer.WriteHeader();
writer.Write(variants);
writer.Dispose();
reader = new VcfReader(outputFilePath);
writtenHeader = reader.HeaderLines;
reader.Dispose();
Assert.Equal(expectedHeader2.Count, writtenHeader.Count);
for (int i = 0; i < expectedHeader2.Count; i++)
{
//let version numbers differ
if (expectedHeader1[i].StartsWith("##VariantPhaser=Scylla"))
{
Assert.True(writtenHeader[i].StartsWith("##VariantPhaser=Scylla"));
continue;
}
Assert.Equal(expectedHeader2[i], writtenHeader[i]);
}
}
public static void DoFiltering(PsaraOptions settings)
{
var geometricFilter = new GeometricFilter(settings.GeometricFilterParameters);
//maybe expand to add other filters..
var vcfIn = settings.InputVcf;
var vcfName = Path.GetFileName(vcfIn);
var outputFile = Path.Combine(settings.OutputDirectory, vcfName.Replace(".vcf", ".filtered.vcf"));
outputFile = outputFile.Replace(".genome.filtered.vcf", ".filtered.genome.vcf");
Logger.WriteToLog("filtering " + vcfIn + "...");
if (File.Exists(outputFile))
{
File.Delete(outputFile);
}
List <string> header = VcfReader.GetAllHeaderLines(vcfIn);
string cmdLine = "##Psara_cmdline=" + settings.QuotedCommandLineArgumentsString;
VcfWriterConfig config = GetWriterConfigToMatchInputVcf(vcfIn);
using (PsaraVcfWriter writer = new PsaraVcfWriter(outputFile, config, new VcfWriterInputContext(), header, cmdLine))
{
writer.WriteHeader();
using (VcfReader reader = new VcfReader(vcfIn, false))
{
var backLogVcfVariant = new VcfVariant();
var coLocatedAlleles = new List <CalledAllele>();
var moreVariantsInVcf = reader.GetNextVariant(backLogVcfVariant);
var incomingBatch = new List <CalledAllele>();
while (moreVariantsInVcf)
{
if (incomingBatch.Count == 0)
{
incomingBatch = moreVariantsInVcf ? VcfVariantUtilities.Convert(new List <VcfVariant> {
backLogVcfVariant
},
config.ShouldOutputRcCounts, config.ShouldOutputTsCounts, false).ToList() : null;
moreVariantsInVcf = reader.GetNextVariant(backLogVcfVariant);
}
if ((coLocatedAlleles.Count == 0) || AreColocated(coLocatedAlleles, incomingBatch))
{
coLocatedAlleles.AddRange(incomingBatch);
incomingBatch.Clear();
//colocated alleles are left behind
}
else
{
FilterAndStreamOut(coLocatedAlleles, writer, geometricFilter);
coLocatedAlleles.Clear();
//incomingBatch alleles are left behind
}
}
//if you get here, there is no more unprocessed vcf variants but there could be
//colocated or an incoming batch of alleles left over. We need to write them to file before exiting.
FilterAndStreamOut(coLocatedAlleles, writer, geometricFilter);
FilterAndStreamOut(incomingBatch, writer, geometricFilter);
}
}
}
public void WriteADiploidNbhd()
{
var outputDir = Path.Combine(TestPaths.LocalScratchDirectory, "MergerWriteADiploidNbhd");
var outputFilePath = Path.Combine(outputDir, "TinyDiploid.Phased.vcf");
var inputFilePath = Path.Combine(TestPaths.LocalTestDataDirectory, "TinyDiploid.vcf");
var expectedFilePath = Path.Combine(TestPaths.LocalTestDataDirectory, "TinyDiploidOutput.vcf");
TestHelper.RecreateDirectory(outputDir);
var context = new VcfWriterInputContext
{
QuotedCommandLineString = "myCommandLine",
SampleName = "mySample",
ReferenceName = "myReference",
ContigsByChr = new List <Tuple <string, long> >
{
new Tuple <string, long>("chr1", 10001),
new Tuple <string, long>("chr22", 51304566),
new Tuple <string, long>("chrX", 500)
}
};
var config = new VcfWriterConfig
{
DepthFilterThreshold = 500,
VariantQualityFilterThreshold = 30,
FrequencyFilterThreshold = 0.007f,
ShouldOutputNoCallFraction = true,
ShouldOutputStrandBiasAndNoiseLevel = true,
EstimatedBaseCallQuality = 23,
PloidyModel = PloidyModel.DiploidByThresholding,
AllowMultipleVcfLinesPerLoci = false
};
var writer = new PhasedVcfWriter(outputFilePath, config, new VcfWriterInputContext(), new List <string>()
{
}, null);
var reader = new AlleleReader(inputFilePath, true);
//set up the original variants
var originalVcfVariant1 = TestHelper.CreateDummyAllele("chr1", 1, "A", "G", 1000, 156);
var originalVcfVariant2 = TestHelper.CreateDummyAllele("chr1", 1, "A", "T", 1000, 156);
var originalVcfVariant4 = TestHelper.CreateDummyAllele("chr22", 1230237, "GTC", "G", 1000, 156);
var originalVcfVariant5 = TestHelper.CreateDummyAllele("chr22", 1230237, "GTC", "GTCT", 1000, 156);
var vs1 = new VariantSite((originalVcfVariant1));
var vs2 = new VariantSite((originalVcfVariant2));
var vs4 = new VariantSite((originalVcfVariant4));
var vs5 = new VariantSite((originalVcfVariant5));
//have to replace variants at positon 116380048 (we call two new MNVS here)
var nbhd1 = new VcfNeighborhood(0, "chr1", vs1, vs2);
var calledNbh1 = new CallableNeighborhood(nbhd1, new VariantCallingParameters());
VcfMerger merger = new VcfMerger(reader);
List <Tuple <CalledAllele, string> > alleleTuplesPastNbhd = new List <Tuple <CalledAllele, string> >();
//we will just say, we called the variants that were in the origina vcf. Ie, we agree with it.
calledNbh1.CalledVariants = new Dictionary <int, List <CalledAllele> > {
{ originalVcfVariant1.ReferencePosition, new List <CalledAllele> {
originalVcfVariant1, originalVcfVariant2
} }
};
//Realizes the first nbhd starts at chr1 . We have to do something with the first lines of the vcf (chr1 1 . A G,T)
//so, alleleTuplesPastNbhd = chr1 1 . A G,T
alleleTuplesPastNbhd = merger.WriteVariantsUptoChr(writer, alleleTuplesPastNbhd, nbhd1.ReferenceName);
Assert.True(alleleTuplesPastNbhd[0].Item1.IsSameAllele(originalVcfVariant1));
Assert.True(alleleTuplesPastNbhd[1].Item1.IsSameAllele(originalVcfVariant2));
//This method writes everything up to the end of nbhd 1,
//so "(chr1 1 . A G,T)" from the vcf and the variants scylla detected "(chr1 1 . A G,T)" need to be dealt with.
//Since these 4 variants are actually the same two, we need to remove the vcf ones and only write the scylla ones.
//Thn we peek into the vcf and see the next line is "chr22 1230237 . GTC G,GTCT", clearly outside nbh1.
//so we write out everything we need for nbhd1, and save the peeked line
alleleTuplesPastNbhd = merger.WriteVariantsUptoIncludingNbhd(writer, alleleTuplesPastNbhd, calledNbh1);
Assert.True(alleleTuplesPastNbhd[0].Item1.IsSameAllele(originalVcfVariant4));
Assert.True(alleleTuplesPastNbhd[1].Item1.IsSameAllele(originalVcfVariant5));
//now write out
//chr22 1230237.GTC G,GTCT 50 DP = 1370 GT: GQ: AD: DP: VF: NL: SB: NC: US 1 / 2:100:185,68:364:0.258:20:-100.0000:0.0000:0,0,0,0,0,0,1,1,0,0,0,2
//chrX 79.CG GTG,AA 50 DP = 1370 GT: GQ: AD: DP: VF: NL: SB: NC: US 1 / 2:100:185,68:364:0.258:20:-100.0000:0.0000:0,0,0,0,0,0,1,1,0,0,0,2
merger.WriteRemainingVariants(writer, alleleTuplesPastNbhd);
writer.Dispose();
var expectedLines = File.ReadLines(expectedFilePath).ToList();
var outputLines = File.ReadLines(outputFilePath).ToList();
Assert.Equal(expectedLines.Count(), outputLines.Count());
for (int i = 0; i < expectedLines.Count; i++)
{
Assert.Equal(expectedLines[i], outputLines[i]);
}
}
public VennVcfFormatter(VcfWriterConfig Config, bool debugMode)
{
_config = Config;
UpdateFrequencyFormat();
DebugMode = debugMode;
}
public static void DoReformating(string inputFile, bool crush)
{
var outputFile = inputFile.Replace(".vcf", ".uncrushed.vcf");
if (crush)
{
Console.WriteLine("crushing " + inputFile + "...");
outputFile = inputFile.Replace(".vcf", ".crushed.vcf");
}
else
{
Console.WriteLine("uncrushing " + inputFile + "...");
}
if (File.Exists(outputFile))
{
File.Delete(outputFile);
}
var config = new VcfWriterConfig()
{
AllowMultipleVcfLinesPerLoci = !crush
};
using (VcfFileWriter writer = new VcfFileWriter(outputFile, config, new VcfWriterInputContext()))
{
writer.WriteHeader();
using (VcfReader reader = new VcfReader(inputFile, false))
{
var currentAllele = new CalledAllele();
var backLogVcfVariant = new VcfVariant();
var backLogExists = reader.GetNextVariant(backLogVcfVariant);
while (backLogExists)
{
var backLogAlleles = backLogExists ? VcfVariantUtilities.Convert(new List <VcfVariant> {
backLogVcfVariant
}).ToList() : null;
foreach (var allele in backLogAlleles)
{
try
{
writer.Write(new List <CalledAllele>()
{
allele
});
}
catch (Exception ex)
{
Console.WriteLine("Problem writing " + allele.ToString());
Console.WriteLine("Exception: " + ex);
return;
}
}
backLogExists = reader.GetNextVariant(backLogVcfVariant);
if (backLogAlleles[0].Chromosome != backLogVcfVariant.ReferenceName)
{
//we have switched to the next chr. flush the buffer.
writer.FlushBuffer();
}
}
writer.FlushBuffer();
}
}
}
