/// <summary>
/// Write out the ploidy bed file if ploidy information is available from the vcf header
/// Only create the normal XX or XY ploidy bed file so that Canvas can properly classify any abnormalities as variant.
/// If ploidy Y is > 1 produce the XY ploidy bed file, otherwise produce the XX ploidy bed file
/// </summary>
public IFileLocation CreateGermlinePloidyBed(Vcf vcf, GenomeMetadata genomeMetadata, IDirectoryLocation sampleSandbox)
{
string sexKaryotype = PloidyCorrector.GetSexChromosomeKaryotypeFromVcfHeader(vcf.VcfFile.FullName);
if (sexKaryotype == null)
{
_logger.Warn($"Sex chromosome ploidy not found in {vcf.VcfFile} header. No ploidy will be provided to Canvas.");
return(null);
}
_logger.Info($"Found sex chromosome ploidy {PloidyCorrector.PrintPloidy(sexKaryotype)} in {vcf.VcfFile}");
var ploidyInfo = new SamplePloidyInfo();
IFileLocation ploidyBed = sampleSandbox.GetFileLocation("ploidy.bed.gz");
if (sexKaryotype.ToLower().Contains("y"))
{
ploidyInfo.ProvidedPloidy = SexPloidyInfo.NormalMale;
_logger.Info($"Creating male ploidy bed file at {ploidyBed}.");
}
else
{
ploidyInfo.ProvidedPloidy = SexPloidyInfo.NormalFemale;
_logger.Info($"Creating female ploidy bed file at {ploidyBed}.");
}
string headerLine = $"{PloidyCorrector.ReferenceSexChromosomeKaryotype}={PloidyCorrector.PrettyPrintPloidy(ploidyInfo.ProvidedPloidyX.Value, ploidyInfo.ProvidedPloidyY.Value)}";
_ploidyFixer.WritePloidyBedFile(ploidyInfo, genomeMetadata, _ploidyFixer.GetParRegions(genomeMetadata),
ploidyBed.FullName, headerLine, ploidy => true);
return(ploidyBed);
}
private CanvasEnrichmentOutput RunSingleSample(string sampleId, CanvasEnrichmentInput input, IFileLocation dbSnpVcf, IFileLocation manifest, IDirectoryLocation sampleSandbox)
{
if (!_annotationFileProvider.IsSupported(input.GenomeMetadata))
{
_logger.Info($"Skipping Canvas for sample {sampleId}: unsupported reference genome '{input.GenomeMetadata.Name}'");
return(null);
}
var moreCustomParameters = new Dictionary <string, string>();
StringBuilder commandLine = new StringBuilder("Somatic-Enrichment");
commandLine.Append(_singleSampleInputCommandLineBuilder.GetSingleSampleCommandLine(sampleId, input.Bam, input.GenomeMetadata, sampleSandbox));
string sexChromosomeKaryotype = null;
if (input.PloidyInfo.IsPloidyAvailable)
{
sexChromosomeKaryotype = input.PloidyInfo.IsXYMale.Value ? PloidyCorrector.PrettyPrintPloidy(1, 1)
: (input.PloidyInfo.IsXXFemale.Value ? PloidyCorrector.PrettyPrintPloidy(2, 0) : null);
}
string controlSexChromosomeKaryotype = null;
if (input.IsCanvasNormalizePcaMode && input.PloidyInfo.IsPloidyAvailable)
{
IFileLocation pcaModelFile = input.PloidyInfo.IsXYMale.Value ? input.PcaModels.MaleModelFile
: (input.PloidyInfo.IsXXFemale.Value ? input.PcaModels.FemaleModelFile : null);
if (pcaModelFile == null)
{
string sampleSex = input.PloidyInfo.IsXYMale.Value ? "male"
: (input.PloidyInfo.IsXXFemale.Value ? "female" : "sex unknown");
_logger.Info($"Skipping Canvas for sample {sampleId}: PCA model file not available for {sampleSex} samples.");
return(null);
}
moreCustomParameters["CanvasNormalize"] = "-m PCA";
commandLine.Append($" --control-binned \"{pcaModelFile}\"");
commandLine.Append($" --control-bin-size 100"); // use a dummy bin size for now
controlSexChromosomeKaryotype = sexChromosomeKaryotype;
}
else if (input.PrecomputedControl?.BinnedPath != null)
{
commandLine.Append($" --control-binned \"{input.PrecomputedControl.BinnedPath}\"");
commandLine.Append($" --control-bin-size {input.PrecomputedControl.BinSize}");
controlSexChromosomeKaryotype = input.PrecomputedControl.SexChromosomeKaryotype;
}
else
{
foreach (var normalBam in input.NormalBamPaths)
{
commandLine.Append($" --control-bam \"{normalBam.BamFile}\"");
}
}
AddControlPloidyVcf(commandLine, input, controlSexChromosomeKaryotype, sampleId, sampleSandbox);
commandLine.Append($" --{SingleSampleCommonOptionsParser.PopulationBAlleleVcfOptionName} \"{dbSnpVcf}\"");
commandLine.Append($" --manifest \"{manifest}\"");
AddSamplePloidyVcf(commandLine, input, sexChromosomeKaryotype, sampleId, sampleSandbox);
if (input.PredefinedBinsFile != null)
{
moreCustomParameters["CanvasBin"] = $"-n={input.PredefinedBinsFile.WrapWithEscapedShellQuote()}";
}
commandLine.Append(_singleSampleInputCommandLineBuilder.GetCustomParameters(moreCustomParameters));
commandLine = _singleSampleInputCommandLineBuilder.MergeCustomCanvasParameters(commandLine);
var job = new JobInfo(_runtimeExecutable.FullName, _canvasExe + " " + commandLine, "Canvas_" + sampleId);
return(_workDoer.DoWork(WorkResourceRequest.CreateExact(1, 8), job, GetCanvasOutput(sampleId, sampleSandbox)).Await());
}
public CanvasPloidyBedCreator(ILogger logger, IWorkManager workManager, PloidyCorrector ploidyFixer)
{
_logger = logger;
_workManager = workManager;
_ploidyFixer = ploidyFixer;
}
private CanvasEnrichmentOutput RunSingleSample(string sampleId, CanvasEnrichmentInput input, IFileLocation dbSnpVcf, IFileLocation manifest, IDirectoryLocation sampleSandbox)
{
if (!_annotationFileProvider.IsSupported(input.GenomeMetadata))
{
_logger.Info($"Skipping Canvas for sample {sampleId}: unsupported reference genome '{input.GenomeMetadata.Name}'");
return(null);
}
var moreCustomParameters = new Dictionary <string, string>();
StringBuilder commandLine = new StringBuilder("Somatic-Enrichment");
commandLine.Append(_singleSampleInputCommandLineBuilder.GetSingleSampleCommandLine(sampleId, input.Bam, input.GenomeMetadata, sampleSandbox));
string sexChromosomeKaryotype = null;
if (input.PloidyInfo.IsPloidyAvailable)
{
sexChromosomeKaryotype = input.PloidyInfo.IsXYMale.Value ? PloidyCorrector.PrettyPrintPloidy(1, 1)
: (input.PloidyInfo.IsXXFemale.Value ? PloidyCorrector.PrettyPrintPloidy(2, 0) : null);
}
string controlSexChromosomeKaryotype = null;
if (input.IsCanvasNormalizePcaMode && input.PloidyInfo.IsPloidyAvailable)
{
IFileLocation pcaModelFile = input.PloidyInfo.IsXYMale.Value ? input.PcaModels.MaleModelFile
: (input.PloidyInfo.IsXXFemale.Value ? input.PcaModels.FemaleModelFile : null);
if (pcaModelFile == null)
{
string sampleSex = input.PloidyInfo.IsXYMale.Value ? "male"
: (input.PloidyInfo.IsXXFemale.Value ? "female" : "sex unknown");
_logger.Info($"Skipping Canvas for sample {sampleId}: PCA model file not available for {sampleSex} samples.");
return(null);
}
moreCustomParameters["CanvasNormalize"] = "-m PCA";
commandLine.Append($" --control-binned \"{pcaModelFile}\"");
commandLine.Append($" --control-bin-size 100"); // use a dummy bin size for now
controlSexChromosomeKaryotype = sexChromosomeKaryotype;
}
else if (input.PrecomputedControl?.BinnedPath != null)
{
commandLine.Append($" --control-binned \"{input.PrecomputedControl.BinnedPath}\"");
commandLine.Append($" --control-bin-size {input.PrecomputedControl.BinSize}");
controlSexChromosomeKaryotype = input.PrecomputedControl.SexChromosomeKaryotype;
}
else
{
foreach (var normalBam in input.NormalBamPaths)
{
commandLine.Append($" --control-bam \"{normalBam.BamFile}\"");
}
}
if (controlSexChromosomeKaryotype != null)
{
var controlPloidyBed = sampleSandbox.GetFileLocation("control-ploidy.bed.gz");
if (_canvasPloidyBedCreator.GeneratePloidyBedFileFromSexChromosomeKaryotype(input.GenomeMetadata, input.GenomeMetadata.Sequences.First().FastaPath,
controlSexChromosomeKaryotype, controlPloidyBed.FullName, sampleSandbox.FullName))
{
commandLine.Append($" --{SomaticEnrichmentOptionsParser.ControlPloidyBedOptionName} \"{controlPloidyBed}\"");
}
}
commandLine.Append($" --{SingleSampleCommonOptionsParser.PopulationBAlleleVcfOptionName} \"{dbSnpVcf}\"");
commandLine.Append($" --manifest \"{manifest}\"");
if (sexChromosomeKaryotype != null)
{
var ploidyBed = sampleSandbox.GetFileLocation("ploidy.bed.gz");
if (_canvasPloidyBedCreator.GeneratePloidyBedFileFromSexChromosomeKaryotype(input.GenomeMetadata, input.GenomeMetadata.Sequences.First().FastaPath,
sexChromosomeKaryotype, ploidyBed.FullName, sampleSandbox.FullName))
{
commandLine.Append($" --{SingleSampleCommonOptionsParser.PloidyBedOptionName} \"{ploidyBed}\"");
}
}
if (input.PredefinedBinsFile != null)
{
moreCustomParameters["CanvasBin"] = $"-n={input.PredefinedBinsFile.WrapWithEscapedShellQuote()}";
}
commandLine.Append(_singleSampleInputCommandLineBuilder.GetCustomParameters(moreCustomParameters));
commandLine = _singleSampleInputCommandLineBuilder.MergeCustomCanvasParameters(commandLine);
UnitOfWork singleSampleJob = new UnitOfWork()
{
ExecutablePath = CrossPlatform.IsThisLinux() ? _runtimeExecutable.FullName : _canvasExe.FullName,
CommandLine = CrossPlatform.IsThisLinux() ? _canvasExe + " " + commandLine : commandLine.ToString(),
LoggingFolder = _workManager.LoggingFolder.FullName,
LoggingStub = "Canvas_" + sampleId,
};
_workManager.DoWorkSingleThread(singleSampleJob);
return(GetCanvasOutput(sampleId, sampleSandbox));
}
public static void Evaluate(string truthSetPath, string cnvCallsPath, string excludedBed, string outputPath, EvaluateCnvOptions options)
{
double heterogeneityFraction = options.HeterogeneityFraction;
var knownCn = LoadKnownCn(truthSetPath, heterogeneityFraction);
knownCn = knownCn.SelectValues(
truthEntries => truthEntries.Where(truthEntry => truthEntry.Length >= options.MinEntrySize).ToList());
var calls = GetCnvCallsFromVcf(cnvCallsPath, options.DQscoreThreshold);
calls = calls.SelectValues(
chromosomeCalls => chromosomeCalls.Where(call => call.Length >= options.MinEntrySize).ToList());
// LoadRegionsOfInterest(options.RoiBed?.FullName);
var excludeIntervals = new Dictionary <string, List <CNInterval> >();
if (!string.IsNullOrEmpty(excludedBed))
{
var excludeIntervalsTmp = LoadIntervalsFromBed(excludedBed, false, 1.0);
List <string> keys = excludeIntervalsTmp.Keys.ToList();
foreach (string key in keys)
{
string chr = key;
if (!calls.ContainsKey(chr))
{
chr = key.Replace("chr", "");
}
if (!calls.ContainsKey(chr))
{
chr = "chr" + key;
}
if (!calls.ContainsKey(chr))
{
Console.WriteLine($"Error: Skipping exclude intervals for chromosome {key} with no truth data." +
$"Check that chromosome names are spelled correctly for exclude intervals");
continue;
}
excludeIntervals[chr] = excludeIntervalsTmp[key];
}
}
Console.WriteLine("TruthSet\t{0}", truthSetPath);
Console.WriteLine("CNVCalls\t{0}", cnvCallsPath);
bool includePassingOnly = Path.GetFileName(cnvCallsPath).ToLower().Contains("vcf");
var logger = new Logger(new[] { Console.Out }, new[] { Console.Error });
var settings = IsasConfigurationSettings.GetConfigSettings();
var output = new DirectoryLocation(outputPath);
var workerDirectory = new DirectoryLocation(Isas.Framework.Utilities.Utilities.GetAssemblyFolder(typeof(CNVChecker)));
var commandManager = new CommandManager(new ExecutableProcessor(settings, logger, workerDirectory));
WorkDoerFactory.RunWithWorkDoer(logger, settings, output, workDoer =>
{
var tabixWrapper = TabixWrapperFactory.GetTabixWrapper(logger, workDoer, commandManager);
var ploidyCorrector = new PloidyCorrector(logger, workDoer,
new PloidyEstimator(logger, workDoer, null, false, commandManager), tabixWrapper, false);
var checker = new CNVChecker(options.DQscoreThreshold, excludeIntervals, ploidyCorrector);
if (options.PloidyInfo.SexPloidyInfo != null)
{
Console.WriteLine($">>>Getting reference ploidy from provided ploidy information and PAR bed file '{options.PloidyInfo.ParBed}'");
var ploidy = checker.GetPloidy(options.PloidyInfo, output);
var referencePloidy = LoadReferencePloidy(options.PloidyInfo.SexPloidyInfo, options.PloidyInfo.ParBed);
knownCn = GetKnownCopyNumberWithReferencePloidy(referencePloidy, knownCn);
calls = GetCallsWithRefPloidy(calls, ploidy);
}
var cnvEvaluator = new CnvEvaluator(checker);
if (checker.DQscoreThreshold.HasValue && !Path.GetFileName(cnvCallsPath).ToLower().Contains("vcf"))
{
throw new ArgumentException("CNV.vcf must be in a vcf format when --dqscore option is used");
}
cnvEvaluator.ComputeAccuracy(knownCn, cnvCallsPath, outputPath, includePassingOnly, options, calls);
if (includePassingOnly)
{
cnvEvaluator.ComputeAccuracy(knownCn, cnvCallsPath, outputPath, false, options, calls);
}
ComputeCallability(logger, calls, options, output);
Console.WriteLine(">>>Done - results written to {0}", outputPath);
});
}
public CanvasPloidyVcfCreator(PloidyCorrector ploidyFixer)
{
_ploidyFixer = ploidyFixer;
}