/// <summary>
/// Integrity check, to ensure that our reference FASTA file is in sync with our inputs.
/// </summary>
private static void SanityCheckChromosomeNames(GenomeMetadata genome, IEnumerable <CanvasSegment> segments)
{
var chromosomeNames = new HashSet <string>();
foreach (GenomeMetadata.SequenceMetadata chromosome in genome.Contigs())
{
chromosomeNames.Add(chromosome.Name.ToLowerInvariant());
}
foreach (
CanvasSegment segment in
segments.Where(segment => !chromosomeNames.Contains(segment.Chr.ToLowerInvariant())))
{
throw new Exception($"Integrity check error: Segment found at unknown chromosome '{segment.Chr}'");
}
}
private static Dictionary <string, IEnumerable <Interval> > GetIncludedIntervals(Dictionary <string, List <Interval> > filterIntervals, GenomeMetadata genomeMetadata)
{
var contigNames = genomeMetadata
.Contigs()
.Where(contig => contig.Type == GenomeMetadata.SequenceType.Autosome || contig.Type == GenomeMetadata.SequenceType.Allosome)
.Select(contig => contig.Name);
return(contigNames
.Select(contig =>
{
var filteredIntervals = filterIntervals.ContainsKey(contig)
? filterIntervals[contig]
: new List <Interval>();
return (contig, GetIncludedIntervals(filteredIntervals, genomeMetadata.GetSequence(contig).Length));
})
.ToDictionary());
}
public static CallabilityMetricsComputer Create(ILogger logger, GenomeMetadata genomeMetadata, IFileLocation filterBed, bool isFemale)
{
var filterIntervals = LoadBedRegions(filterBed, genomeMetadata);
var nonFilterIntervals = GetIncludedIntervals(filterIntervals, genomeMetadata);
string chrY = genomeMetadata
.Contigs()
.Select(contig => contig.Name)
.SingleOrDefault(name => name == "chrY" || name == "Y");
if (isFemale && chrY != null)
{
nonFilterIntervals.Remove(chrY);
}
var callabilityCalculator = new CallabilityCalculator(nonFilterIntervals);
return(new CallabilityMetricsComputer(logger, callabilityCalculator));
}
public StringBuilder GetSingleSampleCommandLine(string sampleId, Bam bam, GenomeMetadata genomeMetadata, IDirectoryLocation sampleSandbox)
{
StringBuilder commandLine = new StringBuilder();
commandLine.Append($" --bam \"{bam.BamFile}\"");
commandLine.Append($" --sample-name \"{sampleId}\"");
IFileLocation kmerFasta = _annotationFileProvider.GetKmerFasta(genomeMetadata);
commandLine.Append($" --reference \"{kmerFasta}\"");
IDirectoryLocation wholeGenomeFasta = new FileLocation(genomeMetadata.Contigs().First().FastaPath).Directory;
commandLine.Append($" --genome-folder \"{wholeGenomeFasta}\"");
IFileLocation filterBed = _annotationFileProvider.GetFilterBed(genomeMetadata);
commandLine.Append($" --filter-bed \"{filterBed}\"");
commandLine.Append($" --output \"{sampleSandbox}\"");
return(commandLine);
}
/// <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 StringBuilder GetMultiSampleCommandLine(SampleSet <CanvasPedigreeSample> samples, GenomeMetadata genomeMetadata, Vcf vcf, IDirectoryLocation sampleSandbox)
{
StringBuilder commandLine = new StringBuilder();
// move proband to the front of collection (enum Proband gets the lowest int value )
var sortedBySampleTypeSamples = samples.OrderBy(sample => sample.Value.SampleType);
foreach (var sampleKvp in sortedBySampleTypeSamples)
{
var sampleId = sampleKvp.Key.Id;
var sample = sampleKvp.Value;
commandLine.Append($" --bam \"{sample.Bam.BamFile}\" {sample.SampleType} {sampleId}");
}
IFileLocation kmerFasta = _annotationFileProvider.GetKmerFasta(genomeMetadata);
commandLine.Append($" --reference \"{kmerFasta}\"");
IDirectoryLocation wholeGenomeFasta = new FileLocation(genomeMetadata.Contigs().First().FastaPath).Directory;
commandLine.Append($" --genome-folder \"{wholeGenomeFasta}\"");
IFileLocation filterBed = _annotationFileProvider.GetFilterBed(genomeMetadata);
commandLine.Append($" --filter-bed \"{filterBed}\"");
commandLine.Append($" --output \"{sampleSandbox}\"");
return(commandLine);
}
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);
}
private IReferenceGenome GetReferenceGenomeFromGenomeMetadata(GenomeMetadata genomeMetadata)
{
return(_referenceGenomeFactory.GetReferenceGenome(new DirectoryLocation(Path.GetDirectoryName(genomeMetadata.Contigs().First().FastaPath))));
}
public bool IsSupported(GenomeMetadata genome)
{
string species = genome.Contigs().FirstOrDefault()?.Species;
return("Homo_sapiens".Equals(species, StringComparison.OrdinalIgnoreCase));
}