private void ExecuteGroupingTest(List <Read> reads, List <int> expectedGroupMemberships, IEnumerable <Tuple <int, string, string> > variants)
{
var variantSites = new List <VariantSite>();
foreach (var variant in variants)
{
variantSites.Add(new VariantSite(variant.Item1)
{
VcfReferenceAllele = variant.Item2, VcfAlternateAllele = variant.Item3
});
}
var alignmentExtractor = new MockAlignmentExtractor(reads);
var veadSource = new VeadGroupSource(alignmentExtractor, new BamFilterParameters()
{
MinimumMapQuality = 20
}, false, "");
var vcfNeighborhood = new VcfNeighborhood(new VariantCallingParameters(), "chr1", new VariantSite(120), new VariantSite(121), "T")
{
VcfVariantSites = variantSites
};
vcfNeighborhood.SetRangeOfInterest();
var veadGroups = veadSource.GetVeadGroups(vcfNeighborhood).ToList();
Assert.Equal(expectedGroupMemberships.Count, veadGroups.Count());
for (var i = 0; i < veadGroups.Count(); i++)
{
Assert.Equal(expectedGroupMemberships[i], veadGroups[i].NumVeads);
}
}
public void GetVeads()
{
var vcfNeighborhood = new VcfNeighborhood(new VariantCallingParameters(), "chr1", new VariantSite(10000), new VariantSite(200000), "T")
{
VcfVariantSites = new List <VariantSite>
{
new VariantSite(100)
{
VcfReferenceAllele = "A", VcfAlternateAllele = "C"
},
new VariantSite(400)
{
VcfReferenceAllele = "A", VcfAlternateAllele = "C"
},
new VariantSite(505)
{
VcfReferenceAllele = "A", VcfAlternateAllele = "C"
},
new VariantSite(703)
{
VcfReferenceAllele = "A", VcfAlternateAllele = "T"
},
new VariantSite(800)
{
VcfReferenceAllele = "A", VcfAlternateAllele = "C"
},
}
};
var reads = new List <Read>();
reads.Add(CreateRead("chr1", "ACGT", 10)); // Before neighborhood
reads.Add(CreateRead("chr1", "ACGT", 96)); // Ends right before neighborhood's first variant site
reads.Add(CreateRead("chr1", "ACGT", 100)); // Match (100)
reads.Add(CreateRead("chr1", "ACGT", 300)); // Within neighborhood but no VariantSite
reads.Add(CreateRead("chr1", "ACGT", 400, qualityForAll: 19)); // Within neighbhorhood but low quals
reads.Add(CreateRead("chr1", "ACGT", 500)); // Within neighborhood but no VariantSite (ends right before 505)
reads.Add(CreateRead("chr1", "ACGT", 700)); // Match (703)
reads.Add(CreateRead("chr1", "ACGT", 800)); // Match (800)
reads.Add(CreateRead("chr1", "ACGT", 805)); // Past neighborhood
reads.Add(CreateRead("chr1", "ACGT", 900)); // Past neighborhood
reads.Add(CreateRead("chr2", "ACGT", 100)); // Wrong chromosome
vcfNeighborhood.SetRangeOfInterest();
var alignmentExtractor = new MockAlignmentExtractor(reads);
var veadSource = new VeadGroupSource(alignmentExtractor,
new BamFilterParameters()
{
MinimumMapQuality = 20
}, false, "");
var veadGroups = veadSource.GetVeadGroups(vcfNeighborhood);
// Collect all reads that could relate to the neighborhood
// - Skip anything that has quality less than MinimumMapQuality
// - Skip anything that ends before neighborhood begins
// - Stop collecting once we've passed the end of the neighborhood
// We should have collected the reads at 100, 700, and 800.
Assert.Equal(801, vcfNeighborhood.LastPositionOfInterestWithLookAhead);
Assert.Equal(3, veadGroups.Count());
Assert.Equal(1, veadGroups.Count(v => v.RepresentativeVead.Name.EndsWith("100")));
Assert.Equal(1, veadGroups.Count(v => v.RepresentativeVead.Name.EndsWith("700")));
Assert.Equal(1, veadGroups.Count(v => v.RepresentativeVead.Name.EndsWith("800")));
Assert.Equal(0, veadGroups.Count(v => v.RepresentativeVead.Name.EndsWith("805")));
Assert.Equal(0, veadGroups.Count(v => v.RepresentativeVead.Name.EndsWith("900")));
foreach (var veadGroup in veadGroups)
{
Assert.Equal(1, veadGroup.NumVeads);
}
vcfNeighborhood.VcfVariantSites.Add(
new VariantSite(790)
{
VcfReferenceAllele = "ACAGTGAAAGACTTGTGAC", VcfAlternateAllele = "C"
});
vcfNeighborhood.SetRangeOfInterest();
Assert.Equal(809, vcfNeighborhood.LastPositionOfInterestWithLookAhead);
alignmentExtractor = new MockAlignmentExtractor(reads);
veadSource = new VeadGroupSource(alignmentExtractor,
new BamFilterParameters()
{
MinimumMapQuality = 20
}, false, "");
veadGroups = veadSource.GetVeadGroups(vcfNeighborhood);
Assert.Equal(3, veadGroups.Count());
Assert.Equal(1, veadGroups.Count(v => v.RepresentativeVead.Name.EndsWith("100")));
Assert.Equal(1, veadGroups.Count(v => v.RepresentativeVead.Name.EndsWith("700")));
Assert.Equal(1, veadGroups.Count(v => v.RepresentativeVead.Name.EndsWith("800")));
Assert.Equal(0, veadGroups.Count(v => v.RepresentativeVead.Name.EndsWith("805")));
Assert.Equal(0, veadGroups.Count(v => v.RepresentativeVead.Name.EndsWith("900")));
// Boundary case - read ends exactly at neighborhood's first variant site
reads = new List <Read>();
reads.Add(CreateRead("chr1", "ACGT", 10)); // Before neighborhood
reads.Add(CreateRead("chr1", "ACGT", 96)); // Ends right before neighborhood's first variant site
reads.Add(CreateRead("chr1", "ACGT", 97)); // Ends exactly at neighborhood's first variant site
alignmentExtractor = new MockAlignmentExtractor(reads);
veadSource = new VeadGroupSource(alignmentExtractor, new BamFilterParameters()
{
MinimumMapQuality = 20
}, false, "");
veadGroups = veadSource.GetVeadGroups(vcfNeighborhood);
// The veadgroup for 97 should be the only one
Assert.Equal(1, veadGroups.Count());
Assert.Equal(1, veadGroups.Count(v => v.RepresentativeVead.Name.EndsWith("97")));
foreach (var veadGroup in veadGroups)
{
Assert.Equal(1, veadGroup.NumVeads);
}
}
