public void GetPloidyFromGenotypes_DotIsIgnored()
{
var genotypes = new[] { Genotype.GetGenotype("."), Genotype.GetGenotype("1|2"), Genotype.GetGenotype("0/2") };
var ploidy = AlleleBlock.GetMaxPloidy(genotypes);
Assert.Equal(2, ploidy);
}
public void GetAlleleBlockToSampleHaplotype_AlleleBlock_WithInternalRefPositions_SplitIfOutOfRange()
{
var genotypeBlock1 = new GenotypeBlock(new[] { "1|2", "0/0", "0|0", "1/1" }.Select(Genotype.GetGenotype).ToArray());
var genotypeBlock2 = new GenotypeBlock(new[] { "1/1", "0|0", "1|1" }.Select(Genotype.GetGenotype).ToArray());
var genotypeBlock3 = new GenotypeBlock(new[] { "1|2", "0|0", "1|1" }.Select(Genotype.GetGenotype).ToArray(), 1);
var genotypeToSample =
new Dictionary <GenotypeBlock, List <int> >
{
{ genotypeBlock1, new List <int> {
0
} },
{ genotypeBlock2, new List <int> {
1
} },
{ genotypeBlock3, new List <int> {
2
} }
};
var indexOfUnsupportedVars = Enumerable.Repeat(new HashSet <int>(), genotypeBlock1.Genotypes.Length).ToArray();
var starts = new[] { 100, 102, 103, 104 };
var functionBlockRanges = starts.Select(x => x + 2).ToList();
var alleleBlockToSampleHaplotype = AlleleBlock.GetAlleleBlockToSampleHaplotype(genotypeToSample,
indexOfUnsupportedVars, starts, functionBlockRanges, out _);
var expectedBlock1 = new AlleleBlock(1, new[] { 1, 0, 1 }, 0, 0);
var expectedBlock2 = new AlleleBlock(1, new[] { 2, 0, 1 }, 0, 0);
Assert.True(alleleBlockToSampleHaplotype.ContainsKey(expectedBlock1));
Assert.True(alleleBlockToSampleHaplotype[expectedBlock1].SequenceEqual(new[] { new SampleHaplotype(2, 0) }));
Assert.True(alleleBlockToSampleHaplotype.ContainsKey(expectedBlock2));
Assert.True(alleleBlockToSampleHaplotype[expectedBlock2].SequenceEqual(new[] { new SampleHaplotype(2, 1) }));
}
public void ExtendAlleleBlock_AsExpected()
{
var alleleBlock1 = new AlleleBlock(2, new [] { 1, 1 }, 2, 2);
var extendedBlock1 = AlleleBlockMerger.ExtendAlleleBlock(alleleBlock1, 2, 2);
var expectedBlock1 = new AlleleBlock(0, new [] { 0, 0, 1, 1, 0, 0 }, -1, -1);
Assert.Equal(extendedBlock1, expectedBlock1);
}
private static VariantInfo GetVariantInfo(PositionSet positionSet, AlleleBlock alleleBlock)
{
var positions = positionSet.SimplePositions;
int startIndex = alleleBlock.PositionIndex;
int numPositions = alleleBlock.AlleleIndexes.Length;
int numSamples = positionSet.NumSamples;
string qual = GetStringWithMinValueOrDot(Enumerable.Range(startIndex, numPositions)
.Select(x => positions[x].VcfFields[VcfCommon.QualIndex]));
var filters = Enumerable.Range(startIndex, numPositions)
.Select(i => positions[i].VcfFields[VcfCommon.FilterIndex])
.ToArray();
var gqValues = new string[numSamples];
for (var i = 0; i < numSamples; i++)
{
gqValues[i] = GetStringWithMinValueOrDot(
new ArraySegment <string>(positionSet.GqInfo.Values[i], startIndex, numPositions).ToArray());
}
var psValues = new string[numSamples];
for (var i = 0; i < numSamples; i++)
{
var psTagsThisSample =
new ArraySegment <string>(positionSet.PsInfo.Values[i], startIndex, numPositions);
var isHomozygous = new ArraySegment <bool>(
positionSet.GtInfo.Values[i].Select(x => x.IsHomozygous).ToArray(), startIndex, numPositions);
psValues[i] = GetPhaseSetForRecomposedVariant(psTagsThisSample, isHomozygous);
}
var homoReferenceSamplePloidy = new int?[numSamples];
for (var i = 0; i < numSamples; i++)
{
if (Genotype.IsAllHomozygousReference(positionSet.GtInfo.Values[i], startIndex, numPositions))
{
homoReferenceSamplePloidy[i] = positionSet.GtInfo.Values[i][startIndex].AlleleIndexes.Length;
}
}
var sampleFilters = new List <bool> [numSamples];
for (var i = 0; i < numSamples; i++)
{
sampleFilters[i] = new List <bool>();
}
return(new VariantInfo(qual, filters, gqValues, psValues, homoReferenceSamplePloidy, sampleFilters));
}
public void GetAlleleBlockToSampleHaplotype_AwareOfTrimmedRefPositions()
{
var genotypeBlock1 = new GenotypeBlock(new[] { "0|0", "1|1", "1|1", "0|0" }.Select(Genotype.GetGenotype).ToArray());
var genotypeBlock2 = new GenotypeBlock(new[] { "0|0", "1|1", "1|1" }.Select(Genotype.GetGenotype).ToArray());
var genotypeBlock3 = new GenotypeBlock(new[] { "1|1", "1|1", "0|0" }.Select(Genotype.GetGenotype).ToArray(), 1);
var genotypeBlock4 = new GenotypeBlock(new[] { "1|1", "1|1" }.Select(Genotype.GetGenotype).ToArray(), 1);
var genotypeToSample =
new Dictionary <GenotypeBlock, List <int> >
{
{ genotypeBlock1, new List <int> {
0
} },
{ genotypeBlock2, new List <int> {
1
} },
{ genotypeBlock3, new List <int> {
2
} },
{ genotypeBlock4, new List <int> {
3
} }
};
var indexOfUnsupportedVars = Enumerable.Repeat(new HashSet <int>(), 4).ToArray();
var starts = Enumerable.Range(100, 4).ToArray();
var functionBlockRanges = starts.Select(x => x + 2).ToList();
var alleleBlockToSampleHaplotype = AlleleBlock.GetAlleleBlockToSampleHaplotype(genotypeToSample,
indexOfUnsupportedVars, starts, functionBlockRanges, out _);
var expectedBlock1 = new AlleleBlock(1, new[] { 1, 1 }, 1, 1);
var expectedBlock2 = new AlleleBlock(1, new[] { 1, 1 }, 1, 0);
var expectedBlock3 = new AlleleBlock(1, new[] { 1, 1 }, 0, 1);
var expectedBlock4 = new AlleleBlock(1, new[] { 1, 1 }, 0, 0);
Assert.True(alleleBlockToSampleHaplotype.ContainsKey(expectedBlock1));
Assert.True(alleleBlockToSampleHaplotype[expectedBlock1]
.SequenceEqual(new[] { new SampleHaplotype(0, 0), new SampleHaplotype(0, 1) }));
Assert.True(alleleBlockToSampleHaplotype.ContainsKey(expectedBlock2));
Assert.True(alleleBlockToSampleHaplotype[expectedBlock2]
.SequenceEqual(new[] { new SampleHaplotype(1, 0), new SampleHaplotype(1, 1) }));
Assert.True(alleleBlockToSampleHaplotype.ContainsKey(expectedBlock3));
Assert.True(alleleBlockToSampleHaplotype[expectedBlock3]
.SequenceEqual(new[] { new SampleHaplotype(2, 0), new SampleHaplotype(2, 1) }));
Assert.True(alleleBlockToSampleHaplotype.ContainsKey(expectedBlock4));
Assert.True(alleleBlockToSampleHaplotype[expectedBlock4]
.SequenceEqual(new[] { new SampleHaplotype(3, 0), new SampleHaplotype(3, 1) }));
}
public void GetAlleleBlockToSampleHaplotype_AlleleBlock_OneAlleleIsRef_EachTime()
{
var genotypeBlock1 = new GenotypeBlock(new[] { "1|0", "0|1", "1|0", "0|1" }.Select(Genotype.GetGenotype).ToArray());
var genotypeBlock2 = new GenotypeBlock(new[] { "1/1", "0|1", "1|0" }.Select(Genotype.GetGenotype).ToArray(), 1);
var genotypeBlock3 = new GenotypeBlock(new[] { "0|0", "1|0", "0|1", "0|0" }.Select(Genotype.GetGenotype).ToArray());
var genotypeBlock4 = new GenotypeBlock(new[] { "0|1", "1|0", "1|0" }.Select(Genotype.GetGenotype).ToArray());
var genotypeToSample =
new Dictionary <GenotypeBlock, List <int> >
{
{ genotypeBlock1, new List <int> {
0
} },
{ genotypeBlock2, new List <int> {
1
} },
{ genotypeBlock3, new List <int> {
2
} },
{ genotypeBlock4, new List <int> {
3
} }
};
var indexOfUnsupportedVars = Enumerable.Repeat(new HashSet <int>(), genotypeBlock1.Genotypes.Length).ToArray();
var starts = new[] { 100, 101, 102, 104 };
var functionBlockRanges = starts.Select(x => x + 2).ToList();
var alleleBlockToSampleHaplotype = AlleleBlock.GetAlleleBlockToSampleHaplotype(genotypeToSample,
indexOfUnsupportedVars, starts, functionBlockRanges, out _);
var expectedBlock1 = new AlleleBlock(0, new[] { 1, 0, 1 }, 0, 0);
var expectedBlock2 = new AlleleBlock(0, new[] { 0, 1, 0 }, 0, 0);
var expectedBlock3 = new AlleleBlock(1, new[] { 1, 0 }, 0, 0);
var expectedBlock4 = new AlleleBlock(1, new[] { 1, 1 }, 0, 0);
var expectedBlock5 = new AlleleBlock(1, new[] { 0, 0 }, 0, 0);
Assert.True(alleleBlockToSampleHaplotype.ContainsKey(expectedBlock1));
Assert.True(alleleBlockToSampleHaplotype[expectedBlock1].SequenceEqual(new[] { new SampleHaplotype(0, 0) }));
Assert.True(alleleBlockToSampleHaplotype.ContainsKey(expectedBlock2));
Assert.True(alleleBlockToSampleHaplotype[expectedBlock2].SequenceEqual(new[] { new SampleHaplotype(0, 1) }));
Assert.True(alleleBlockToSampleHaplotype.ContainsKey(expectedBlock3));
Assert.True(alleleBlockToSampleHaplotype[expectedBlock3].SequenceEqual(new[] { new SampleHaplotype(1, 0) }));
Assert.True(alleleBlockToSampleHaplotype.ContainsKey(expectedBlock4));
Assert.True(alleleBlockToSampleHaplotype[expectedBlock4].SequenceEqual(new[] { new SampleHaplotype(1, 1), new SampleHaplotype(3, 0) }));
Assert.True(alleleBlockToSampleHaplotype.ContainsKey(expectedBlock5));
Assert.True(alleleBlockToSampleHaplotype[expectedBlock5].SequenceEqual(new[] { new SampleHaplotype(3, 1) }));
}
public void GetPositionsAndRefAltAlleles_AsExpected()
{
var genotypeBlock = new GenotypeBlock(new[] { "1|2", "1/1", "0|1", "0/1" }.Select(Genotype.GetGenotype).ToArray());
var genotypeToSample =
new Dictionary <GenotypeBlock, List <int> > {
{ genotypeBlock, new List <int> {
0
} }
};
var indexOfUnsupportedVars = Enumerable.Repeat(new HashSet <int>(), genotypeBlock.Genotypes.Length).ToArray();
var starts = new[] { 356, 358, 360, 361 };
var functionBlockRanges = new List <int> {
358, 360, 362, 364
};
var alleles = new[] { new[] { "G", "C", "T" }, new[] { "A", "T" }, new[] { "C", "G" }, new[] { "G", "T" } };
const string refSequence = "GAATCG";
var alleleBlockToSampleHaplotype = AlleleBlock.GetAlleleBlockToSampleHaplotype(genotypeToSample, indexOfUnsupportedVars, starts, functionBlockRanges, out var alleleBlockGraph);
var mergedAlleleBlockToSampleHaplotype =
AlleleBlockMerger.Merge(alleleBlockToSampleHaplotype, alleleBlockGraph).ToArray();
var alleleSet = new AlleleSet(ChromosomeUtilities.Chr1, starts, alleles);
var alleleBlocks = mergedAlleleBlockToSampleHaplotype.Select(x => x.Key).ToArray();
var sequence = new NSequence();
var result1 = VariantGenerator.GetPositionsAndRefAltAlleles(alleleBlocks[0], alleleSet, refSequence, starts[0], null, sequence, _vidCreator);
var result2 = VariantGenerator.GetPositionsAndRefAltAlleles(alleleBlocks[1], alleleSet, refSequence, starts[0], null, sequence, _vidCreator);
var expectedVarPosIndexes1 = new List <int> {
0, 1
};
var expectedVarPosIndexes2 = new List <int> {
0, 1, 2
};
Assert.Equal((356, 360, "GAATC", "CATTC"), (result1.Start, result1.End, result1.Ref, result1.Alt));
for (var i = 0; i < expectedVarPosIndexes1.Count; i++)
{
Assert.Equal(expectedVarPosIndexes1[i], result1.VarPosIndexesInAlleleBlock[i]);
}
Assert.Equal((356, 360, "GAATC", "TATTG"), (result2.Start, result2.End, result2.Ref, result2.Alt));
for (var i = 0; i < expectedVarPosIndexes2.Count; i++)
{
Assert.Equal(expectedVarPosIndexes2[i], result2.VarPosIndexesInAlleleBlock[i]);
}
}
decomposedVids) GetPositionsAndRefAltAlleles(AlleleBlock alleleBlock, AlleleSet alleleSet,
string totalRefSequence, int regionStart, List <ISimplePosition> simplePositions, ISequence sequence, IVariantIdCreator vidCreator)
{
int numPositions = alleleBlock.AlleleIndexes.Length;
int firstPositionIndex = alleleBlock.PositionIndex;
int lastPositionIndex = alleleBlock.PositionIndex + numPositions - 1;
int blockStart = alleleSet.Starts[firstPositionIndex];
int blockEnd = alleleSet.Starts[lastPositionIndex];
string lastRefAllele = alleleSet.VariantArrays[lastPositionIndex][0];
int blockRefLength = blockEnd - blockStart + lastRefAllele.Length;
string refSequence = totalRefSequence.Substring(blockStart - regionStart, blockRefLength);
var refSequenceStart = 0;
var altSequenceSegments = new LinkedList <string>();
var variantPosIndexesInAlleleBlock = new List <int>();
var vidListsNeedUpdate = new List <List <string> >();
var decomposedVids = new List <string>();
if (FindConflictAllele(alleleBlock, alleleSet))
{
return(default);
public void Merge_AsExpected()
{
var genotypeBlock1 = new GenotypeBlock(new[] { "1|0", "1|1", "1|1" }.Select(Genotype.GetGenotype).ToArray());
var genotypeBlock2 = new GenotypeBlock(new[] { "0|0", "1|0", "1|1" }.Select(Genotype.GetGenotype).ToArray());
var genotypeToSample =
new Dictionary <GenotypeBlock, List <int> >
{
{ genotypeBlock1, new List <int> {
0
} },
{ genotypeBlock2, new List <int> {
1
} }
};
var indexOfUnsupportedVars = Enumerable.Repeat(new HashSet <int>(), 3).ToArray();
var starts = Enumerable.Range(100, 3).ToArray();
var functionBlockRanges = starts.Select(x => x + 2).ToList();
var alleleBlockToSampleHaplotype = AlleleBlock.GetAlleleBlockToSampleHaplotype(genotypeToSample,
indexOfUnsupportedVars, starts, functionBlockRanges, out var alleleBlockGraph);
var mergedAlleleBlockToSampleHaplotype =
AlleleBlockMerger.Merge(alleleBlockToSampleHaplotype, alleleBlockGraph);
var expectedBlock1 = new AlleleBlock(0, new[] { 1, 1, 1 }, -1, -1);
var expectedBlock2 = new AlleleBlock(0, new[] { 0, 1, 1 }, -1, -1);
var expectedBlock3 = new AlleleBlock(0, new[] { 0, 0, 1 }, -1, -1);
Assert.True(mergedAlleleBlockToSampleHaplotype.ContainsKey(expectedBlock1));
Assert.True(mergedAlleleBlockToSampleHaplotype[expectedBlock1]
.SequenceEqual(new[] { new SampleHaplotype(0, 0) }));
Assert.True(mergedAlleleBlockToSampleHaplotype.ContainsKey(expectedBlock2));
Assert.True(mergedAlleleBlockToSampleHaplotype[expectedBlock2]
.SequenceEqual(new[] { new SampleHaplotype(0, 1), new SampleHaplotype(1, 0) }));
Assert.True(mergedAlleleBlockToSampleHaplotype.ContainsKey(expectedBlock3));
Assert.True(mergedAlleleBlockToSampleHaplotype[expectedBlock3]
.SequenceEqual(new[] { new SampleHaplotype(1, 1) }));
}
internal static (int Start, int End, string Ref, string Alt, List <int> VarPosIndexesInAlleleBlock, List <string> decomposedVids) GetPositionsAndRefAltAlleles(AlleleBlock alleleBlock, AlleleSet alleleSet, string totalRefSequence, int regionStart, List <ISimplePosition> simplePositions)
{
int numPositions = alleleBlock.AlleleIndexes.Length;
int firstPositionIndex = alleleBlock.PositionIndex;
int lastPositionIndex = alleleBlock.PositionIndex + numPositions - 1;
int blockStart = alleleSet.Starts[firstPositionIndex];
int blockEnd = alleleSet.Starts[lastPositionIndex];
string lastRefAllele = alleleSet.VariantArrays[lastPositionIndex][0];
int blockRefLength = blockEnd - blockStart + lastRefAllele.Length;
var refSequence = totalRefSequence.Substring(blockStart - regionStart, blockRefLength);
int refSequenceStart = 0;
var altSequenceSegments = new LinkedList <string>();
var variantPosIndexesInAlleleBlock = new List <int>();
var vidListsNeedUpdate = new List <List <string> >();
var decomposedVids = new List <string>();
for (int positionIndex = firstPositionIndex; positionIndex <= lastPositionIndex; positionIndex++)
{
int indexInBlock = positionIndex - firstPositionIndex;
int alleleIndex = alleleBlock.AlleleIndexes[indexInBlock];
//only non-reference alleles considered
if (alleleIndex == 0)
{
continue;
}
variantPosIndexesInAlleleBlock.Add(positionIndex - firstPositionIndex);
string refAllele = alleleSet.VariantArrays[positionIndex][0];
string altAllele = alleleSet.VariantArrays[positionIndex][alleleIndex];
int positionOnRefSequence = alleleSet.Starts[positionIndex] - blockStart;
int refRegionBetweenTwoAltAlleles = positionOnRefSequence - refSequenceStart;
if (refRegionBetweenTwoAltAlleles < 0)
{
string previousAltAllele = alleleSet.VariantArrays[positionIndex - 1][alleleIndex];
throw new UserErrorException($"Conflicting alternative alleles identified at {alleleSet.Chromosome.UcscName}:{alleleSet.Starts[positionIndex]}: both \"{previousAltAllele}\" and \"{altAllele}\" are present.");
}
string refSequenceBefore = refSequence.Substring(refSequenceStart, refRegionBetweenTwoAltAlleles);
altSequenceSegments.AddLast(refSequenceBefore);
altSequenceSegments.AddLast(altAllele);
refSequenceStart = positionOnRefSequence + refAllele.Length;
if (simplePositions == null)
{
continue;
}
var thisPosition = simplePositions[positionIndex];
// alleleIndex is 1-based for altAlleles
int varIndex = alleleIndex - 1;
//Only SNVs get recomposed for now
if (thisPosition.Vids[varIndex] == null)
{
thisPosition.Vids[varIndex] = SmallVariantCreator.GetVid(alleleSet.Chromosome.EnsemblName,
thisPosition.Start, thisPosition.End, thisPosition.AltAlleles[varIndex], VariantType.SNV);
thisPosition.IsDecomposed[varIndex] = true;
}
decomposedVids.Add(thisPosition.Vids[varIndex]);
if (thisPosition.LinkedVids[varIndex] == null)
{
thisPosition.LinkedVids[varIndex] = new List <string>();
}
vidListsNeedUpdate.Add(thisPosition.LinkedVids[varIndex]);
}
altSequenceSegments.AddLast(refSequence.Substring(refSequenceStart));
var recomposedAllele = string.Concat(altSequenceSegments);
var blockRefEnd = blockStart + blockRefLength - 1;
var recomposedVariantId = SmallVariantCreator.GetVid(alleleSet.Chromosome.EnsemblName, blockStart, blockRefEnd, recomposedAllele, VariantType.MNV);
vidListsNeedUpdate.ForEach(x => x.Add(recomposedVariantId));
return(blockStart, blockRefEnd, refSequence, recomposedAllele, variantPosIndexesInAlleleBlock, decomposedVids);
}
