/// <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($";CIEND={segment.EndConfidenceInterval.Item1},{segment.EndConfidenceInterval.Item2}");
}
}
/// <summary>
/// Write to a file a single CanvasSegment record as a non-sample VCF columns
/// </summary>
/// <param name="writer"></param>
/// <param name="firstSampleSegment"></param>
/// <param name="alternateAllele"></param>
/// <param name="recordLevelFilter"></param>
/// <param name="sampleSetCnvType"></param>
/// <param name="isMultisample"></param>
/// <returns></returns>
private static void WriteColumnsUntilInfoField(BgzipOrStreamWriter writer, CanvasSegment firstSampleSegment, CnvType sampleSetCnvType, string alternateAllele, string recordLevelFilter, bool isMultisample)
{
// From vcf 4.1 spec:
// If any of the ALT alleles is a symbolic allele (an angle-bracketed ID String <ID>) then the padding base is required and POS denotes the
// coordinate of the base preceding the polymorphism.
int position = (alternateAllele.StartsWith("<") && alternateAllele.EndsWith(">"))
? firstSampleSegment.Begin
: firstSampleSegment.Begin + 1;
writer.Write($"{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($";CIEND={firstSampleSegment.EndConfidenceInterval.Item1},{firstSampleSegment.EndConfidenceInterval.Item2}");
}
}
/// <summary>
/// Outputs the copy number calls to a text file.
/// </summary>
public static void WriteSegments(string outVcfPath, List <CanvasSegment> segments, string wholeGenomeFastaDirectory, string sampleName,
List <string> extraHeaders, PloidyInfo ploidy, int qualityThreshold = 10)
{
using (BgzipOrStreamWriter writer = new BgzipOrStreamWriter(outVcfPath))
{
// Write the VCF header:
writer.WriteLine("##fileformat=VCFv4.1");
writer.WriteLine($"##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();
}
}
}
}