/// <summary>
/// Outputs the copy number calls to a text file.
/// </summary>
private static void WriteVariants(IReadOnlyCollection <List <CanvasSegment> > segments, PloidyInfo ploidy, GenomeMetadata genome,
BgzipOrStreamWriter writer, bool isPedigreeInfoSupplied = true, int?denovoQualityThreshold = null)
{
var nSamples = segments.Count;
foreach (GenomeMetadata.SequenceMetadata chromosome in genome.Sequences)
{
for (int segmentIndex = 0; segmentIndex < segments.First().Count; segmentIndex++)
{
var firstSampleSegment = segments.First()[segmentIndex];
if (!isPedigreeInfoSupplied && segments.Select(sample => sample[segmentIndex].Filter == "PASS").Any() && segments.Count > 1)
{
firstSampleSegment.Filter = "PASS";
}
if (!firstSampleSegment.Chr.Equals(chromosome.Name, StringComparison.OrdinalIgnoreCase))
{
continue;
}
var referenceCopyNumbers = segments.Select(segment => ploidy?.GetReferenceCopyNumber(segment[segmentIndex]) ?? 2).ToList();
var currentSegments = segments.Select(x => x[segmentIndex]).ToList();
var cnvTypes = new List <CnvType>();
for (int sampleIndex = 0; sampleIndex < nSamples; sampleIndex++)
{
cnvTypes.Add(currentSegments[sampleIndex].GetCnvType(referenceCopyNumbers[sampleIndex]));
}
CnvType cnvType;
if (cnvTypes.TrueForAll(x => x == CnvType.Reference))
{
cnvType = CnvType.Reference;
}
else if (cnvTypes.TrueForAll(x => x == CnvType.Reference | x == CnvType.Loss))
{
cnvType = CnvType.Loss;
}
else if (cnvTypes.TrueForAll(x => x == CnvType.Reference | x == CnvType.Gain))
{
cnvType = CnvType.Gain;
}
else if (cnvTypes.TrueForAll(x => x == CnvType.Reference | x == CnvType.LossOfHeterozygosity))
{
cnvType = CnvType.LossOfHeterozygosity;
}
else
{
cnvType = CnvType.ComplexCnv;
}
WriteInfoField(writer, firstSampleSegment, cnvType, denovoQualityThreshold, isMultisample: segments.Count > 1);
// FORMAT field
if (segments.Count == 1)
{
WriteSingleSampleInfo(writer, firstSampleSegment);
}
else
{
WriteFormatField(writer, currentSegments);
}
}
}
}
private static void WriteFormatAndSampleFields(BgzipOrStreamWriter writer, CanvasSegment[] segments, string[] genotypes, bool reportDQ)
{
const string nullValue = ".";
string formatColumn = "GT:RC:BC:CN:MCC:MCCQ:QS:FT";
if (reportDQ)
{
formatColumn += ":DQ";
}
var outputFields = new List <string> {
formatColumn
};
for (int i = 0; i < segments.Length; i++)
{
var segment = segments[i];
string mcc = segment.MajorChromosomeCount.HasValue ? segment.MajorChromosomeCount.ToString() : nullValue;
string mccq = segment.MajorChromosomeCountScore.HasValue ? $"{segment.MajorChromosomeCountScore.Value:F2}" : nullValue;
string sampleColumn = $"{genotypes[i]}:{segment.MedianCount:F2}:{segment.BinCount}:{segment.CopyNumber}:{mcc}:{mccq}:{segment.QScore:F2}:{segment.Filter.ToVcfString()}";
if (reportDQ)
{
string dqscore = segment.DqScore.HasValue ? $"{segment.DqScore.Value:F2}" : nullValue;
sampleColumn += $":{dqscore}";
}
outputFields.Add(sampleColumn);
}
writer.WriteLine("\t" + string.Join("\t", outputFields));
}
public static void WriteMultiSampleSegments(string outVcfPath, ISampleMap <List <CanvasSegment> > segments, List <double> diploidCoverage,
string wholeGenomeFastaDirectory, List <string> sampleNames, List <string> extraHeaders, List <PloidyInfo> ploidies, int qualityThreshold, int?denovoQualityThreshold, int?sizeThreshold, bool isPedigreeInfoSupplied = true)
{
using (BgzipOrStreamWriter writer = new BgzipOrStreamWriter(outVcfPath))
{
var genome = WriteVcfHeader(segments.Values.First(), diploidCoverage.Average(), wholeGenomeFastaDirectory, sampleNames,
extraHeaders, writer, qualityThreshold, denovoQualityThreshold, sizeThreshold);
WriteVariants(segments.Zip(), ploidies, genome, writer, denovoQualityThreshold);
}
}
public static void WriteMultiSampleSegments(string outVcfPath, List <List <CanvasSegment> > segments, List <double?> diploidCoverage,
string wholeGenomeFastaDirectory, List <string> sampleNames, List <string> extraHeaders, PloidyInfo ploidy, int qualityThreshold,
bool isPedigreeInfoSupplied = true, int?denovoQualityThreshold = null)
{
using (BgzipOrStreamWriter writer = new BgzipOrStreamWriter(outVcfPath))
{
var genome = WriteVcfHeader(segments.First(), GetMean(diploidCoverage), wholeGenomeFastaDirectory, sampleNames,
extraHeaders, qualityThreshold, writer, denovoQualityThreshold);
WriteVariants(segments, ploidy, genome, writer, isPedigreeInfoSupplied, denovoQualityThreshold);
}
}
public static void WriteHeaderAllAltCnTags(BgzipOrStreamWriter writer, int maxCopyNum = 5)
{
foreach (var copyNum in Enumerable.Range(0, maxCopyNum + 1))
{
if (copyNum == 1)
{
continue;
}
writer.WriteLine($"##ALT=<ID=CN{copyNum},Description=\"Copy number allele: {copyNum} copies\">");
}
}
public void Write(IEnumerable <BedGraphEntry> bedGraphEntries, IFileLocation location, string header = null)
{
using (var bgzipWriter = new BgzipOrStreamWriter(location.FullName))
{
if (header != null)
{
bgzipWriter.WriteLine(header);
}
var writer = new BedGraphWriter(bgzipWriter);
writer.WriteLines(bedGraphEntries);
}
}
private static GenomeMetadata WriteVcfHeader(List <CanvasSegment> segments, double?diploidCoverage,
string wholeGenomeFastaDirectory, List <string> sampleNames, List <string> extraHeaders, int qualityThreshold,
BgzipOrStreamWriter writer, int?denovoQualityThreshold = null)
{
// Write the VCF header:
writer.WriteLine("##fileformat=VCFv4.1");
writer.WriteLine($"##source={CanvasVersionInfo.NameString} {CanvasVersionInfo.VersionString}");
writer.WriteLine($"##reference={Path.Combine(wholeGenomeFastaDirectory, "genome.fa")}");
AddPloidyAndCoverageHeaders(writer, segments, diploidCoverage);
foreach (string header in extraHeaders ?? new List <string>())
{
writer.WriteLine(header);
}
GenomeMetadata genome = new GenomeMetadata();
genome.Deserialize(Path.Combine(wholeGenomeFastaDirectory, "GenomeSize.xml"));
foreach (GenomeMetadata.SequenceMetadata chromosome in genome.Sequences)
{
writer.WriteLine($"##contig=<ID={chromosome.Name},length={chromosome.Length}>");
}
string qualityFilter = $"q{qualityThreshold}";
writer.WriteLine("##ALT=<ID=CNV,Description=\"Copy number variable region\">");
writer.WriteLine($"##FILTER=<ID={qualityFilter},Description=\"Quality below {qualityThreshold}\">");
writer.WriteLine("##FILTER=<ID=L10kb,Description=\"Length shorter than 10kb\">");
writer.WriteLine("##INFO=<ID=CIEND,Number=2,Type=Integer,Description=\"Confidence interval around END for imprecise variants\">");
writer.WriteLine("##INFO=<ID=CIPOS,Number=2,Type=Integer,Description=\"Confidence interval around POS for imprecise variants\">");
writer.WriteLine("##INFO=<ID=CNVLEN,Number=1,Type=Integer,Description=\"Number of reference positions spanned by this CNV\">");
writer.WriteLine("##INFO=<ID=END,Number=1,Type=Integer,Description=\"End position of the variant described in this record\">");
writer.WriteLine("##INFO=<ID=SVTYPE,Number=1,Type=String,Description=\"Type of structural variant\">");
writer.WriteLine("##INFO=<ID=SUBCLONAL,Number=0,Type=Flag,Description=\"Subclonal variant\">");
if (denovoQualityThreshold.HasValue)
{
string denovoQualityFilter = $"dq{denovoQualityThreshold}";
writer.WriteLine($"##INFO=<ID={denovoQualityFilter},Description=\"De novo quality score above {denovoQualityThreshold.Value}\">");
}
writer.WriteLine("##FORMAT=<ID=RC,Number=1,Type=Float,Description=\"Mean counts per bin in the region\">");
writer.WriteLine("##FORMAT=<ID=BC,Number=1,Type=Float,Description=\"Number of bins in the region\">");
writer.WriteLine("##FORMAT=<ID=CN,Number=1,Type=Integer,Description=\"Copy number genotype for imprecise events\">");
writer.WriteLine("##FORMAT=<ID=MCC,Number=1,Type=Integer,Description=\"Major chromosome count (equal to copy number for LOH regions)\">");
if (denovoQualityThreshold.HasValue)
{
writer.WriteLine("##FORMAT=<ID=DQ,Number=1,Type=Float,Description=\"De novo variants Phred-scaled quality score\">");
writer.WriteLine("##FORMAT=<ID=QS,Number=1,Type=Float,Description=\"Phred-scaled quality score\">");
}
string names = string.Join("\t", sampleNames.ToArray());
writer.WriteLine("#CHROM\tPOS\tID\tREF\tALT\tQUAL\tFILTER\tINFO\tFORMAT\t" + names);
SanityCheckChromosomeNames(genome, segments);
return(genome);
}
/// <summary>
/// Write to a file a single CanvasSegment record as a non-sample VCF columns
/// </summary>
/// <param name="writer"></param>
/// <param name="segment"></param>
/// <param name="cnvType"></param>
/// <param name="denovoQualityThreshold"></param>
/// <returns></returns>
private static void WriteInfoField(BgzipOrStreamWriter writer, CanvasSegment segment, CnvType cnvType, int?denovoQualityThreshold, bool isMultisample)
{
// From vcf 4.1 spec:
// If any of the ALT alleles is a symbolic allele (an angle-bracketed ID String “<ID>”) then the padding base is required and POS denotes the
// coordinate of the base preceding the polymorphism.
string alternateAllele = cnvType.ToAltId();
int position = (alternateAllele.StartsWith("<") && alternateAllele.EndsWith(">"))
? segment.Begin
: segment.Begin + 1;
writer.Write($"{segment.Chr}\t{position}\tCanvas:{cnvType.ToVcfId()}:{segment.Chr}:{segment.Begin + 1}-{segment.End}\t");
string qScore = "";
qScore = isMultisample ? "." : $"{segment.QScore:F2}";
writer.Write($"N\t{alternateAllele}\t{qScore}\t{segment.Filter}\t");
if (cnvType != CnvType.Reference)
{
writer.Write($"SVTYPE={cnvType.ToSvType()};");
}
if (segment.IsHeterogeneous)
{
writer.Write("SUBCLONAL;");
}
if (segment.DQScore.HasValue && !isMultisample)
{
writer.Write($"DQ={segment.DQScore.Value};");
}
if (denovoQualityThreshold.HasValue & segment.DQScore.HasValue & segment.DQScore >= denovoQualityThreshold)
{
writer.Write($"dq{denovoQualityThreshold};");
}
writer.Write($"END={segment.End}");
if (cnvType != CnvType.Reference)
{
writer.Write($";CNVLEN={segment.End - segment.Begin}");
}
if (segment.StartConfidenceInterval != null)
{
writer.Write($";CIPOS={segment.StartConfidenceInterval.Item1},{segment.StartConfidenceInterval.Item2}");
}
if (segment.EndConfidenceInterval != null)
{
writer.Write($";CIEND={segment.EndConfidenceInterval.Item1},{segment.EndConfidenceInterval.Item2}");
}
}
private static void WriteFormatField(BgzipOrStreamWriter writer, List <CanvasSegment> segments)
{
writer.Write("\tRC:BC:CN:MCC:QS:DQ");
const string nullValue = ".";
foreach (var segment in segments)
{
var mcc = segment.MajorChromosomeCount.HasValue ? segment.MajorChromosomeCount.ToString() : nullValue;
var dqscore = segment.DQScore.HasValue ? $"{segment.DQScore.Value:F2}" : nullValue;
var rc = Math.Round(segment.MeanCount, 0, MidpointRounding.AwayFromZero);
writer.Write($"\t{rc}:{segment.BinCount}:{ segment.CopyNumber}:{mcc}:{segment.QScore}:{dqscore}");
}
writer.WriteLine();
}
public static void WriteSegments(string outVcfPath, List <CanvasSegment> segments, double?diploidCoverage,
string wholeGenomeFastaDirectory, string sampleName,
List <string> extraHeaders, PloidyInfo ploidy, int qualityThreshold, bool isPedigreeInfoSupplied, int?denovoQualityThreshold = null)
{
using (BgzipOrStreamWriter writer = new BgzipOrStreamWriter(outVcfPath))
{
var genome = WriteVcfHeader(segments, diploidCoverage, wholeGenomeFastaDirectory, new List <string> {
sampleName
},
extraHeaders, qualityThreshold, writer, denovoQualityThreshold);
WriteVariants(new List <List <CanvasSegment> > {
segments.ToList()
}, ploidy, genome, writer, isPedigreeInfoSupplied, denovoQualityThreshold);
}
}
private static void WriteSingleSampleInfo(BgzipOrStreamWriter writer, CanvasSegment segment)
{
writer.Write("\tRC:BC:CN", segment.End);
if (segment.MajorChromosomeCount.HasValue)
{
writer.Write(":MCC");
}
writer.Write("\t{1}:{2}:{3}", segment.End, Math.Round(segment.MeanCount, 0, MidpointRounding.AwayFromZero),
segment.BinCount, segment.CopyNumber);
if (segment.MajorChromosomeCount.HasValue)
{
writer.Write(":{0}", segment.MajorChromosomeCount);
}
writer.WriteLine();
}
/// <summary>
/// Write to a file a single CanvasSegment record as a non-sample VCF columns
/// </summary>
/// <param name="writer"></param>
/// <param name="firstSampleSegment"></param>
/// <param name="alternateAllele"></param>
/// <param name="recordLevelFilter"></param>
/// <param name="sampleSetCnvType"></param>
/// <param name="isMultisample"></param>
/// <returns></returns>
private static void WriteColumnsUntilInfoField(BgzipOrStreamWriter writer, CanvasSegment firstSampleSegment, CnvType sampleSetCnvType, string alternateAllele, string recordLevelFilter, bool isMultisample)
{
// From vcf 4.1 spec:
// If any of the ALT alleles is a symbolic allele (an angle-bracketed ID String <ID>) then the padding base is required and POS denotes the
// coordinate of the base preceding the polymorphism.
int position = (alternateAllele.StartsWith("<") && alternateAllele.EndsWith(">"))
? firstSampleSegment.Begin
: firstSampleSegment.Begin + 1;
writer.Write($"{firstSampleSegment.Chr}\t{position}\tCanvas:{sampleSetCnvType.ToVcfId()}:{firstSampleSegment.Chr}:{firstSampleSegment.Begin + 1}-{firstSampleSegment.End}\t");
string qScore = isMultisample ? "." : $"{firstSampleSegment.QScore:F2}";
writer.Write($"N\t{alternateAllele}\t{qScore}\t{recordLevelFilter}\t");
if (sampleSetCnvType != CnvType.Reference)
{
writer.Write($"SVTYPE={sampleSetCnvType.ToSvType()};");
}
if (firstSampleSegment.IsHeterogeneous)
{
writer.Write("SUBCLONAL;");
}
if (firstSampleSegment.IsCommonCnv)
{
writer.Write("COMMONCNV;");
}
writer.Write($"END={firstSampleSegment.End}");
if (sampleSetCnvType != CnvType.Reference)
{
writer.Write($";CNVLEN={firstSampleSegment.Length}");
}
if (firstSampleSegment.StartConfidenceInterval != null)
{
writer.Write($";CIPOS={firstSampleSegment.StartConfidenceInterval.Item1},{firstSampleSegment.StartConfidenceInterval.Item2}");
}
if (firstSampleSegment.EndConfidenceInterval != null)
{
writer.Write($";CIEND={firstSampleSegment.EndConfidenceInterval.Item1},{firstSampleSegment.EndConfidenceInterval.Item2}");
}
}
public static void WriteSegments(string outVcfPath, List <CanvasSegment> segments, double?diploidCoverage,
string wholeGenomeFastaDirectory, string sampleName,
List <string> extraHeaders, PloidyInfo ploidy, int qualityThreshold, bool isPedigreeInfoSupplied, int?denovoQualityThreshold, int?sizeThreshold)
{
using (BgzipOrStreamWriter writer = new BgzipOrStreamWriter(outVcfPath))
{
var genome = WriteVcfHeader(segments, diploidCoverage, wholeGenomeFastaDirectory, new List <string> {
sampleName
},
extraHeaders, writer, qualityThreshold, denovoQualityThreshold, sizeThreshold);
var sampleId = new SampleId(sampleName);
var segmentsOfAllSamples = segments.Select(x => new SampleMap <CanvasSegment> {
{ sampleId, x }
});
WriteVariants(segmentsOfAllSamples, new List <PloidyInfo> {
ploidy
}, genome, writer, denovoQualityThreshold);
}
}
public static void AddPloidyAndCoverageHeaders(BgzipOrStreamWriter writer, List <CanvasSegment> segments, double?diploidCoverage)
{
double totalPloidy = 0;
double totalWeight = 0;
foreach (CanvasSegment segment in segments.Where(segment => segment.Filter.IsPass))
{
totalWeight += segment.Length;
totalPloidy += segment.CopyNumber * (segment.Length);
}
if (totalWeight > 0)
{
writer.WriteLine($"##OverallPloidy={totalPloidy / totalWeight:F2}");
if (diploidCoverage != null)
{
writer.WriteLine($"##DiploidCoverage={diploidCoverage:F2}");
}
}
}
/// <summary>
/// Outputs the copy number calls to a text file.
/// </summary>
private static void WriteVariants(IEnumerable <ISampleMap <CanvasSegment> > segmentsOfAllSamples, List <PloidyInfo> ploidies, GenomeMetadata genome,
BgzipOrStreamWriter writer, int?denovoQualityThreshold = null)
{
var segmentsOfAllSamplesArray = segmentsOfAllSamples.ToArray(); // TODO: not necessary when chrom match logic has been updated
int nSamples = segmentsOfAllSamplesArray.First().Values.Count();
foreach (GenomeMetadata.SequenceMetadata chromosome in genome.Contigs()) //TODO: this is extremely inefficient. Segments should be sorted by chromosome
{
foreach (var sampleMap in segmentsOfAllSamplesArray)
{
var currentSegments = sampleMap.Values.ToArray();
var firstSampleSegment = currentSegments.First();
if (!firstSampleSegment.Chr.Equals(chromosome.Name, StringComparison.OrdinalIgnoreCase)
) //TODO: this is extremely inefficient. Segments should be sorted by chromosome
{
continue;
}
var recordLevelFilter = CanvasFilter.GetRecordLevelFilterFromSampleFiltersOnly(
sampleMap
.Select(x => x.Value.Filter)
.ToReadOnlyList())
.ToVcfString();
var referenceCopyNumbers = currentSegments.Zip(ploidies,
(segment, ploidy) => ploidy?.GetReferenceCopyNumber(segment) ?? 2).ToList();
var cnvTypes = new CnvType[nSamples];
var sampleSetAlleleCopyNumbers = new int[nSamples][];
for (int sampleIndex = 0; sampleIndex < nSamples; sampleIndex++)
{
(cnvTypes[sampleIndex], sampleSetAlleleCopyNumbers[sampleIndex]) = currentSegments[sampleIndex]
.GetCnvTypeAndAlleleCopyNumbers(referenceCopyNumbers[sampleIndex]);
}
var sampleSetCnvType = AssignCnvType(cnvTypes);
var(alternateAllele, genotypes) = GetAltAllelesAndGenotypes(sampleSetAlleleCopyNumbers);
WriteColumnsUntilInfoField(writer, firstSampleSegment, sampleSetCnvType, alternateAllele,
recordLevelFilter, nSamples > 1);
WriteFormatAndSampleFields(writer, currentSegments, genotypes,
denovoQualityThreshold.HasValue);
}
}
}
public void Write(BgzipOrStreamWriter w, int nSamples)
{
SanityCheck(nSamples);
WriteStringFieldWithTab(w, ChromName);
w.WriteTab(Pos.ToString());
WriteStringFieldWithTab(w, Id);
WriteStringFieldWithTab(w, Ref);
WriteStringFieldWithTab(w, GetAltString());
WriteStringFieldWithTab(w, _qual);
WriteStringFieldWithTab(w, GetFilterString());
WriteStringFieldWithTab(w, GetInfoString());
WriteStringFieldWithTab(w, GetFormatString());
for (int sampleIndex = 0; sampleIndex < nSamples; ++sampleIndex)
{
if (sampleIndex == (nSamples - 1))
WriteStringFieldNoTab(w, GetSampleString(sampleIndex));
else
WriteStringFieldWithTab(w, GetSampleString(sampleIndex));
}
w.WriteLine();
}
public VCFOutStreamer(BgzipOrStreamWriter writer)
: this(writer, "SAMPLE", "SAMPLE")
{
}
private void WriteStringFieldNoTab(BgzipOrStreamWriter w, string s)
{
w.Write(StringCheck(s));
}
public VCFOutStreamer(BgzipOrStreamWriter writer, string[] sampleNames, string[] sampleIds)
{
_sampleNames = sampleNames;
_sampleIDs = sampleIds;
_writer = writer;
_headerList = new List<KeyValuePair<string, string>>();
}
public VCFOutStreamer(BgzipOrStreamWriter writer, string sampleName, string sampleId)
: this(writer, new[] { sampleName }, new[] { sampleId })
{
}
private static GenomeMetadata WriteVcfHeader(List <CanvasSegment> segments, double?diploidCoverage,
string wholeGenomeFastaDirectory, List <string> sampleNames, List <string> extraHeaders, BgzipOrStreamWriter writer, int qualityThreshold,
int?denovoQualityThreshold, int?sizeThreshold)
{
// Write the VCF header:
writer.WriteLine("##fileformat=VCFv4.1");
writer.WriteLine($"##source={CanvasVersionInfo.NameString} {CanvasVersionInfo.VersionString}");
writer.WriteLine($"##reference={Path.Combine(wholeGenomeFastaDirectory, "genome.fa")}");
// Write ##OverallPloidy and ##DiploidCoverage for a single-sample file (where it makes sense to do so):
if (sampleNames.Count == 1)
{
AddPloidyAndCoverageHeaders(writer, segments, diploidCoverage);
}
foreach (string header in extraHeaders ?? new List <string>())
{
writer.WriteLine(header);
}
GenomeMetadata genome = new GenomeMetadata();
genome.Deserialize(new FileLocation(Path.Combine(wholeGenomeFastaDirectory, "GenomeSize.xml")));
foreach (GenomeMetadata.SequenceMetadata chromosome in genome.Contigs())
{
writer.WriteLine($"##contig=<ID={chromosome.Name},length={chromosome.Length}>");
}
string qualityFilter = $"q{qualityThreshold}";
writer.WriteLine("##ALT=<ID=DUP,Description=\"Region of elevated copy number relative to the reference\">");
WriteHeaderAllAltCnTags(writer);
writer.WriteLine($"##FILTER=<ID={qualityFilter},Description=\"Quality below {qualityThreshold}\">");
if (sizeThreshold.HasValue)
{
string sizeFilterName = CanvasFilter.GetCnvSizeFilter(sizeThreshold.Value, out var sizeFilterThreshold);
writer.WriteLine($"##FILTER=<ID={sizeFilterName},Description=\"Length shorter than {sizeFilterThreshold.Number} {sizeFilterThreshold.Units}\">");
}
writer.WriteLine("##FILTER=<ID=FailedFT,Description=\"Sample-level filter failed in all the samples\">");
writer.WriteLine("##INFO=<ID=CIEND,Number=2,Type=Integer,Description=\"Confidence interval around END for imprecise variants\">");
writer.WriteLine("##INFO=<ID=CIPOS,Number=2,Type=Integer,Description=\"Confidence interval around POS for imprecise variants\">");
writer.WriteLine("##INFO=<ID=CNVLEN,Number=1,Type=Integer,Description=\"Number of reference positions spanned by this CNV\">");
writer.WriteLine("##INFO=<ID=END,Number=1,Type=Integer,Description=\"End position of the variant described in this record\">");
writer.WriteLine("##INFO=<ID=SVTYPE,Number=1,Type=String,Description=\"Type of structural variant\">");
writer.WriteLine("##INFO=<ID=SUBCLONAL,Number=0,Type=Flag,Description=\"Subclonal variant\">");
writer.WriteLine("##INFO=<ID=COMMONCNV,Number=0,Type=Flag,Description=\"Common CNV variant identified from pre-specified bed intervals\">");
writer.WriteLine("##FORMAT=<ID=GT,Number=1,Type=String,Description=\"Genotype\">");
writer.WriteLine("##FORMAT=<ID=RC,Number=1,Type=Float,Description=\"Mean counts per bin in the region\">");
writer.WriteLine("##FORMAT=<ID=BC,Number=1,Type=Float,Description=\"Number of bins in the region\">");
writer.WriteLine("##FORMAT=<ID=CN,Number=1,Type=Integer,Description=\"Copy number genotype for imprecise events\">");
writer.WriteLine("##FORMAT=<ID=MCC,Number=1,Type=Integer,Description=\"Major chromosome count (equal to copy number for LOH regions)\">");
writer.WriteLine("##FORMAT=<ID=MCCQ,Number=1,Type=Float,Description=\"Major chromosome count quality score\">");
writer.WriteLine("##FORMAT=<ID=QS,Number=1,Type=Float,Description=\"Phred-scaled quality score. If CN is reference then this is -10log10(prob(variant)) otherwise this is -10log10(prob(no variant).\">");
if (denovoQualityThreshold.HasValue)
{
writer.WriteLine($"##FORMAT=<ID=DQ,Number=1,Type=Float,Description=\"De novo quality. Threshold for passing de novo call: {denovoQualityThreshold}\">");
}
writer.WriteLine("##FORMAT=<ID=FT,Number=1,Type=String,Description=\"Sample filter, 'PASS' indicates that all filters have passed for this sample\">");
var titleColumns = new List <string> {
"#CHROM", "POS", "ID", "REF", "ALT", "QUAL", "FILTER", "INFO", "FORMAT"
};
titleColumns.AddRange(sampleNames);
writer.WriteLine(string.Join("\t", titleColumns));
SanityCheckChromosomeNames(genome, segments);
return(genome);
}
/// <summary>
/// Outputs the copy number calls to a text file.
/// </summary>
public static void WriteSegments(string outVcfPath, List <CanvasSegment> segments, string wholeGenomeFastaDirectory, string sampleName,
List <string> extraHeaders, PloidyInfo ploidy, int qualityThreshold = 10)
{
using (BgzipOrStreamWriter writer = new BgzipOrStreamWriter(outVcfPath))
{
// Write the VCF header:
writer.WriteLine("##fileformat=VCFv4.1");
writer.WriteLine($"##source={CanvasVersionInfo.NameString} {CanvasVersionInfo.VersionString}");
writer.WriteLine($"##reference={Path.Combine(wholeGenomeFastaDirectory, "genome.fa")}");
foreach (string header in extraHeaders ?? new List <string>())
{
writer.WriteLine(header);
}
GenomeMetadata genome = new GenomeMetadata();
genome.Deserialize(Path.Combine(wholeGenomeFastaDirectory, "GenomeSize.xml"));
foreach (GenomeMetadata.SequenceMetadata chromosome in genome.Sequences)
{
writer.WriteLine($"##contig=<ID={chromosome.Name},length={chromosome.Length}>");
}
string qualityFilter = $"q{qualityThreshold}";
writer.WriteLine("##ALT=<ID=CNV,Description=\"Copy number variable region\">");
writer.WriteLine($"##FILTER=<ID={qualityFilter},Description=\"Quality below {qualityThreshold}\">");
writer.WriteLine("##FILTER=<ID=L10kb,Description=\"Length shorter than 10kb\">");
writer.WriteLine("##INFO=<ID=SVTYPE,Number=1,Type=String,Description=\"Type of structural variant\">");
writer.WriteLine("##INFO=<ID=END,Number=1,Type=Integer,Description=\"End position of the variant described in this record\">");
writer.WriteLine("##INFO=<ID=CNVLEN,Number=1,Type=Integer,Description=\"Number of reference positions spanned by this CNV\">");
writer.WriteLine("##FORMAT=<ID=RC,Number=1,Type=Float,Description=\"Mean counts per bin in the region\">");
writer.WriteLine("##FORMAT=<ID=BC,Number=1,Type=Float,Description=\"Number of bins in the region\">");
writer.WriteLine("##FORMAT=<ID=CN,Number=1,Type=Integer,Description=\"Copy number genotype for imprecise events\">");
writer.WriteLine("##FORMAT=<ID=MCC,Number=1,Type=Integer,Description=\"Major chromosome count (equal to copy number for LOH regions)\">");
writer.WriteLine("#CHROM\tPOS\tID\tREF\tALT\tQUAL\tFILTER\tINFO\tFORMAT\t" + sampleName);
SanityCheckChromosomeNames(genome, segments);
foreach (GenomeMetadata.SequenceMetadata chromosome in genome.Sequences)
{
foreach (CanvasSegment segment in segments)
{
if (!segment.Chr.Equals(chromosome.Name, StringComparison.OrdinalIgnoreCase))
{
continue;
}
int referenceCopyNumber = ploidy?.GetReferenceCopyNumber(segment) ?? 2;
CnvType cnvType = segment.GetCnvType(referenceCopyNumber);
// From vcf 4.1 spec:
// If any of the ALT alleles is a symbolic allele (an angle-bracketed ID String “<ID>”) then the padding base is required and POS denotes the
// coordinate of the base preceding the polymorphism.
string alternateAllele = cnvType.ToAltId();
int position = (alternateAllele.StartsWith("<") && alternateAllele.EndsWith(">")) ? segment.Begin : segment.Begin + 1;
writer.Write($"{segment.Chr}\t{position}\tCanvas:{cnvType.ToVcfId()}:{segment.Chr}:{segment.Begin + 1}-{segment.End}\t");
writer.Write($"N\t{alternateAllele}\t{segment.QScore}\t{segment.Filter}\t", alternateAllele, segment.QScore, segment.Filter);
if (cnvType != CnvType.Reference)
{
writer.Write($"SVTYPE={cnvType.ToSvType()};");
}
writer.Write($"END={segment.End}");
if (cnvType != CnvType.Reference)
{
writer.Write($";CNVLEN={segment.End - segment.Begin}");
}
// FORMAT field
writer.Write("\tRC:BC:CN", segment.End);
if (segment.MajorChromosomeCount.HasValue)
{
writer.Write(":MCC");
}
writer.Write("\t{1}:{2}:{3}", segment.End, Math.Round(segment.MeanCount, 0, MidpointRounding.AwayFromZero), segment.BinCount, segment.CopyNumber);
if (segment.MajorChromosomeCount.HasValue)
{
writer.Write(":{0}", segment.MajorChromosomeCount);
}
writer.WriteLine();
}
}
}
}
