public void GetAlignmentSet()
{
var reads = new List <Read>();
reads.Add(CreateRead("chr1", "ACGT", 10, "read1", matePosition: 10));
reads.Add(CreateRead("chr1", "ACGT", 10, "read2", matePosition: 10));
reads.Add(CreateRead("chr1", "ACGT", 10, "read1", read2: true, matePosition: 10)); // mate
reads.Add(CreateRead("chr1", "ACGT", 10, "read_notmapped", isMapped: false, isProperPair: false, matePosition: 10));
reads.Add(CreateRead("chr1", "ACGT", 10, "read3", isProperPair: false, read2: true, matePosition: 10)); // mate
reads.Add(CreateRead("chr1", "ACGT", 10, "read2", read2: true, matePosition: 10));
var extractor = new MockAlignmentExtractor(reads);
var config = new AlignmentSourceConfig {
MinimumMapQuality = 10, OnlyUseProperPairs = false
};
var source = new AlignmentSource(extractor, null, config);
Read read;
var numSets = 0;
while ((read = source.GetNextRead()) != null)
{
numSets++;
Assert.False(read.Name == "read_notmapped");
}
Assert.Equal(numSets, 5);
}
private void ExecuteTest_ProcessReadsSeparately(bool doMate, bool doStitch)
{
var read = CreateRead("chr1", "ACGT", 10, "read1", matePosition: 100);
var readMate = CreateRead("chr1", "ACGT", 100, "read1", read2: true, matePosition: 10);
var extractor = new MockAlignmentExtractor(new List <Read>()
{
read
});
var mateFinder = new Mock <IAlignmentMateFinder>();
mateFinder.Setup(f => f.GetMate(It.IsAny <Read>())).Returns(readMate);
var stitcher = new Mock <IAlignmentStitcher>();
var config = new AlignmentSourceConfig {
MinimumMapQuality = 10, OnlyUseProperPairs = false
};
var source = new AlignmentSource(extractor, doMate ? mateFinder.Object : null, config);
var alignmentSet = source.GetNextRead();
// TODO what do we really want to test here now that we're not pairing?
}
private ISomaticVariantCaller CreateMockVariantCaller(VcfFileWriter vcfWriter, ApplicationOptions options, ChrReference chrRef, MockAlignmentExtractor mae, IStrandBiasFileWriter biasFileWriter = null, string intervalFilePath = null)
{
var config = new AlignmentSourceConfig
{
MinimumMapQuality = options.MinimumMapQuality,
OnlyUseProperPairs = options.OnlyUseProperPairs,
};
IAlignmentStitcher stitcher = null;
if (options.StitchReads)
{
if (options.UseXCStitcher)
{
stitcher = new XCStitcher(options.MinimumBaseCallQuality);
}
else
{
stitcher = new BasicStitcher(options.MinimumBaseCallQuality);
}
}
var mateFinder = options.StitchReads ? new AlignmentMateFinder(MAX_FRAGMENT_SIZE) : null;
var RegionPadder = new RegionPadder(chrRef, null);
var alignmentSource = new AlignmentSource(mae, mateFinder, stitcher, config);
var variantFinder = new CandidateVariantFinder(options.MinimumBaseCallQuality, options.MaxSizeMNV, options.MaxGapBetweenMNV, options.CallMNVs);
var alleleCaller = new AlleleCaller(new VariantCallerConfig
{
IncludeReferenceCalls = options.OutputgVCFFiles,
MinVariantQscore = options.MinimumVariantQScore,
MaxVariantQscore = options.MaximumVariantQScore,
VariantQscoreFilterThreshold = options.FilteredVariantQScore > options.MinimumVariantQScore ? options.FilteredVariantQScore : (int?)null,
MinCoverage = options.MinimumCoverage,
MinFrequency = options.MinimumFrequency,
EstimatedBaseCallQuality = options.AppliedNoiseLevel == -1 ? options.MinimumBaseCallQuality : options.AppliedNoiseLevel,
StrandBiasModel = options.StrandBiasModel,
StrandBiasFilterThreshold = options.StrandBiasAcceptanceCriteria,
FilterSingleStrandVariants = options.FilterOutVariantsPresentOnlyOneStrand,
GenotypeModel = options.GTModel
});
var stateManager = new RegionStateManager();
return(new SomaticVariantCaller(
alignmentSource,
variantFinder,
alleleCaller,
vcfWriter,
stateManager,
chrRef,
RegionPadder,
biasFileWriter));
}
protected virtual IAlignmentSource CreateAlignmentSource(ChrReference chrReference, string bamFilePath)
{
var alignmentExtractor = new BamFileAlignmentExtractor(bamFilePath, _options.StitchReads, chrReference.Name);
var mateFinder = _options.StitchReads ? new AlignmentMateFinder(Constants.MaxFragmentSize) : null; // jg todo - do we want to expose this to command line?
var stitcher = _options.StitchReads ? CreateStitcher() : null;
var config = new AlignmentSourceConfig
{
MinimumMapQuality = _options.MinimumMapQuality,
OnlyUseProperPairs = _options.OnlyUseProperPairs,
};
return(new AlignmentSource(alignmentExtractor, mateFinder, stitcher, config));
}
public void GetAlignmentSet()
{
var reads = new List <Read>();
reads.Add(CreateRead("chr1", "ACGT", 1, "read1"));
reads.Add(CreateRead("chr1", "ACGT", 1, "read2"));
reads.Add(CreateRead("chr1", "ACGT", 1, "read1_mate"));
reads.Add(CreateRead("chr1", "ACGT", 1, "read3_unmapped", isMapped: false, isProperPair: false));
reads.Add(CreateRead("chr1", "ACGT", 1, "read3_mate", isProperPair: false));
reads.Add(CreateRead("chr1", "ACGT", 1, "read2_mate"));
var extractor = new MockAlignmentExtractor(reads);
var mateFinder = new Mock <IAlignmentMateFinder>();
mateFinder.Setup(f => f.GetMate(It.IsAny <Read>())).Returns(
(Read r) =>
r.Name.EndsWith("_mate")
? reads.FirstOrDefault(o => o.Name == r.Name.Replace("_mate", string.Empty))
: null);
var stitcher = new Mock <IAlignmentStitcher>();
stitcher.Setup(s => s.TryStitch(It.IsAny <AlignmentSet>())).Callback((AlignmentSet s) => s.IsStitched = true);
var config = new AlignmentSourceConfig {
MinimumMapQuality = 10, OnlyUseProperPairs = false
};
var source = new AlignmentSource(extractor, mateFinder.Object, stitcher.Object, config);
AlignmentSet set;
var numSets = 0;
while ((set = source.GetNextAlignmentSet()) != null)
{
numSets++;
if (!set.IsFullPair)
{
Assert.True(set.PartnerRead1.Name.Equals("read3_mate"));
Assert.True(set.PartnerRead2 == null);
Assert.False(set.IsStitched);
}
else
{
Assert.True(set.PartnerRead2.Name.Equals(set.PartnerRead1.Name + "_mate"));
Assert.True(set.IsStitched);
}
}
Assert.Equal(numSets, 3);
}
public static ISomaticVariantCaller CreateMockVariantCaller(VcfFileWriter vcfWriter, ApplicationOptions options, ChrReference chrRef, MockAlignmentExtractor mockAlignmentExtractor, IStrandBiasFileWriter biasFileWriter = null, string intervalFilePath = null)
{
var config = new AlignmentSourceConfig
{
MinimumMapQuality = options.MinimumMapQuality,
OnlyUseProperPairs = options.OnlyUseProperPairs,
};
//var mateFinder = options.StitchReads ? new AlignmentMateFinder() : null;
AlignmentMateFinder mateFinder = null;
var alignmentSource = new AlignmentSource(mockAlignmentExtractor, mateFinder, config);
var variantFinder = new CandidateVariantFinder(options.MinimumBaseCallQuality, options.MaxSizeMNV, options.MaxGapBetweenMNV, options.CallMNVs);
var coverageCalculator = new CoverageCalculator();
var alleleCaller = new AlleleCaller(new VariantCallerConfig
{
IncludeReferenceCalls = options.OutputgVCFFiles,
MinVariantQscore = options.MinimumVariantQScore,
MaxVariantQscore = options.MaximumVariantQScore,
VariantQscoreFilterThreshold = options.FilteredVariantQScore > options.MinimumVariantQScore ? options.FilteredVariantQScore : (int?)null,
MinCoverage = options.MinimumDepth,
MinFrequency = options.MinimumFrequency,
EstimatedBaseCallQuality = options.AppliedNoiseLevel == -1 ? options.MinimumBaseCallQuality : options.AppliedNoiseLevel,
StrandBiasModel = options.StrandBiasModel,
StrandBiasFilterThreshold = options.StrandBiasAcceptanceCriteria,
FilterSingleStrandVariants = options.FilterOutVariantsPresentOnlyOneStrand,
ChrReference = chrRef
},
coverageCalculator: coverageCalculator,
variantCollapser: options.Collapse ? new VariantCollapser(null, coverageCalculator) : null);
var stateManager = new RegionStateManager(
expectStitchedReads: mockAlignmentExtractor.SourceIsStitched,
trackOpenEnded: options.Collapse, trackReadSummaries: options.CoverageMethod == CoverageMethod.Approximate);
//statmanager is an allele source
Assert.Equal(0, stateManager.GetAlleleCount(1, AlleleType.A, DirectionType.Forward));
return(new SomaticVariantCaller(
alignmentSource,
variantFinder,
alleleCaller,
vcfWriter,
stateManager,
chrRef,
null,
biasFileWriter));
}
private void ExecuteTest_ProcessReadsSeparately(bool doMate, bool doStitch)
{
var read = CreateRead("chr1", "ACGT", 1, "read1");
var readMate = CreateRead("chr1", "ACGT", 10, "read1mate");
var extractor = new MockAlignmentExtractor(new List <Read>()
{
read
});
var mateFinder = new Mock <IAlignmentMateFinder>();
mateFinder.Setup(f => f.GetMate(It.IsAny <Read>())).Returns(readMate);
var stitcher = new Mock <IAlignmentStitcher>();
var config = new AlignmentSourceConfig {
MinimumMapQuality = 10, OnlyUseProperPairs = false
};
var source = new AlignmentSource(extractor, doMate ? mateFinder.Object : null, doStitch ? stitcher.Object : null, config);
var alignmentSet = source.GetNextAlignmentSet();
// verify stitching relies on pairing
stitcher.Verify(s => s.TryStitch(It.IsAny <AlignmentSet>()), Times.Never);
if (!doMate)
{
TestHelper.CompareReads(read, alignmentSet.PartnerRead1);
Assert.Equal(null, alignmentSet.PartnerRead2);
Assert.Equal(1, alignmentSet.ReadsForProcessing.Count);
TestHelper.CompareReads(read, alignmentSet.ReadsForProcessing.First());
}
else
{
TestHelper.CompareReads(read, alignmentSet.PartnerRead1);
TestHelper.CompareReads(readMate, alignmentSet.PartnerRead2);
Assert.Equal(2, alignmentSet.ReadsForProcessing.Count);
TestHelper.CompareReads(read, alignmentSet.ReadsForProcessing.First());
TestHelper.CompareReads(readMate, alignmentSet.ReadsForProcessing.Last());
}
}
protected virtual IAlignmentSource CreateAlignmentSource(ChrReference chrReference, string bamFilePath, List <string> chrsToProcess = null)
{
AlignmentMateFinder mateFinder = null;
var alignmentExtractor = new BamFileAlignmentExtractor(bamFilePath, chrReference.Name, _bamIntervalLookup.ContainsKey(bamFilePath) && _options.SkipNonIntervalAlignments ? _bamIntervalLookup[bamFilePath] : null);
//Warn if the bam has sequences ordered differently to the reference genome.
//That would confuse us because we will not know how the user wants to order the output gvcf.
if (alignmentExtractor.SequenceOrderingIsNotConsistent(chrsToProcess))
{
Logger.WriteToLog("Warning: Reference sequences in the bam do not match the order of the reference sequences in the genome. Check bam " + bamFilePath);
Logger.WriteToLog("Variants will be ordered according to the reference genome");
}
var config = new AlignmentSourceConfig
{
MinimumMapQuality = _options.MinimumMapQuality,
OnlyUseProperPairs = _options.OnlyUseProperPairs,
};
return(new AlignmentSource(alignmentExtractor, mateFinder, config));
}
protected virtual IAlignmentSource CreateAlignmentSource(ChrReference chrReference, string bamFilePath, bool commandLineSaysStitched, List <string> chrsToProcess = null)
{
AlignmentMateFinder mateFinder = null;
var alignmentExtractor = new BamFileAlignmentExtractor(bamFilePath, commandLineSaysStitched, chrReference.Name);
//Warn if the bam has sequences ordered differently to the reference genome.
//That would confuse us because we will not know how the user wants to order the output gvcf.
if (alignmentExtractor.SequenceOrderingIsNotConsistent(chrsToProcess))
{
Logger.WriteToLog("Warning: Reference sequences in the bam do not match the order of the reference sequences in the genome. Check bam " + bamFilePath);
Logger.WriteToLog("Variants will be ordered according to the reference genome");
}
var config = new AlignmentSourceConfig
{
MinimumMapQuality = _options.BamFilterParameters.MinimumMapQuality,
OnlyUseProperPairs = _options.BamFilterParameters.OnlyUseProperPairs,
SkipDuplicates = _options.BamFilterParameters.RemoveDuplicates
};
return(new AlignmentSource(alignmentExtractor, mateFinder, config));
}
public void Filtering()
{
//bool isMapped = true, bool isPrimaryAlignment = true, bool isProperPair = true, bool isDuplicate = false
var reads = new List <Read>();
reads.Add(CreateRead("chr1", "AAA", 1, "filtered", false, true, true, false, 10));
reads.Add(CreateRead("chr1", "AAA", 1, "filtered", true, false, true, false, 10));
reads.Add(CreateRead("chr1", "AAA", 1, "filtered_properpair", true, true, false, false, 10));
reads.Add(CreateRead("chr1", "AAA", 1, "filtered", true, true, true, true, 10));
reads.Add(CreateRead("chr1", "AAA", 1, "filtered_quality", true, true, true, false, 9));
reads.Add(CreateRead("chr1", "AAA", 1, "Filtered_CigarData", true, true, true, false, 10, false));
reads.Add(CreateRead("chr1", "AAA", 1, "yay", true, true, true, false, 10));
var extractor = new MockAlignmentExtractor(reads);
var config = new AlignmentSourceConfig {
MinimumMapQuality = 10,
OnlyUseProperPairs = true, SkipDuplicates = true
};
var source = new AlignmentSource(extractor, null, config);
Read read;
var fetchedCount = 0;
while ((read = source.GetNextRead()) != null)
{
fetchedCount++;
Assert.Equal(read.Name, "yay");
}
Assert.Equal(1, fetchedCount);
// -------------------------
// turn off proper pairs
config.OnlyUseProperPairs = false;
fetchedCount = 0;
extractor.Reset();
source = new AlignmentSource(extractor, null, config);
while ((read = source.GetNextRead()) != null)
{
fetchedCount++;
Assert.Equal(read.Name,
fetchedCount == 1 ? "filtered_properpair" : "yay");
}
Assert.Equal(fetchedCount, 2);
// -------------------------
// change quality cut off
config.OnlyUseProperPairs = true;
config.MinimumMapQuality = 9;
fetchedCount = 0;
extractor.Reset();
source = new AlignmentSource(extractor, null, config);
while ((read = source.GetNextRead()) != null)
{
fetchedCount++;
Assert.Equal(read.Name,
fetchedCount == 1 ? "filtered_quality" : "yay");
}
Assert.Equal(fetchedCount, 2);
}
