// Need to filter VCF by FILTER = PASS; there are several reasons they don't accept calls that I trust
// There's an attribute "ZYG" for zygosity, either "het" or "h**o" for heterozygous or homozygous
public List <string> CallIndels(string spritzDirectory, int threads, string genomeFastaP, string bedPath, string bamPath, string outdir)
{
CheckInstallation(spritzDirectory);
var vcftools = new VcfToolsWrapper();
IndelVcfPath = Path.Combine(outdir, "variants.indel.vcf");
//IndelVcf1IndexedPath = Path.Combine(outdir, "variants.indel.1index.vcf");
var commands = new List <string>
{
WrapperUtility.ChangeToToolsDirectoryCommand(spritzDirectory),
"if [[ ! -f " + WrapperUtility.ConvertWindowsPath(IndelVcfPath) + " || ! -s " + WrapperUtility.ConvertWindowsPath(IndelVcfPath) + " ]]; then ",
" scalpel-" + ScalpelVersion + "/scalpel-discovery --single " +
"--bam " + WrapperUtility.ConvertWindowsPath(bamPath) +
" --ref " + WrapperUtility.ConvertWindowsPath(genomeFastaP) +
" --bed " + WrapperUtility.ConvertWindowsPath(bedPath) +
" --numprocs " + threads.ToString() +
" --dir " + WrapperUtility.ConvertWindowsPath(outdir),
// scalpel uses 0-indexing, where SnpEff uses 1-indexing, so change this output to match snpeff
//" awk 'BEGIN{OFS=\"\t\"}{ if (substr($0, 1, 1) != \"#\") $2=++$2; print $0 }' " + WrapperUtility.ConvertWindowsPath(IndelVcfPath) + " > " + WrapperUtility.ConvertWindowsPath(IndelVcf1IndexedPath),
"fi",
// vcf-concat doesn't keep all INFO header lines, so just dump the INFO from each variant
vcftools.RemoveAllSnvs(spritzDirectory, IndelVcfPath, false, true)
};
FilteredIndelVcfPath = vcftools.VcfWithoutSnvsPath;
return(commands);
}
/// <summary>
/// HaplotypeCaller for calling variants on each RNA-Seq BAM file individually.
/// </summary>
/// <param name="spritzDirectory"></param>
/// <param name="threads"></param>
/// <param name="genomeFasta"></param>
/// <param name="splitTrimBam"></param>
/// <param name="dbsnpReferenceVcfPath"></param>
/// <param name="newVcf"></param>
public List <string> VariantCalling(string spritzDirectory, ExperimentType experimentType, int threads, string genomeFasta, string splitTrimBam, string dbsnpReferenceVcfPath)
{
HaplotypeCallerGvcfPath = Path.Combine(Path.GetDirectoryName(splitTrimBam), Path.GetFileNameWithoutExtension(splitTrimBam) + ".g.vcf.gz");
HaplotypeCallerVcfPath = Path.Combine(Path.GetDirectoryName(splitTrimBam), Path.GetFileNameWithoutExtension(splitTrimBam) + ".g.gt.vcf");
FilteredHaplotypeCallerVcfPath = Path.Combine(Path.GetDirectoryName(splitTrimBam), Path.GetFileNameWithoutExtension(splitTrimBam) + ".g.gt.NoIndels.vcf");
var vcftools = new VcfToolsWrapper();
List <string> commands = new List <string>
{
WrapperUtility.ChangeToToolsDirectoryCommand(spritzDirectory),
SamtoolsWrapper.GenomeFastaIndexCommand(genomeFasta),
GenomeDictionaryIndexCommand(genomeFasta),
// check that reference VCF is indexed
"if [ ! -f " + WrapperUtility.ConvertWindowsPath(dbsnpReferenceVcfPath) + ".idx ]; then " + Gatk(Workers) + " IndexFeatureFile -F " + WrapperUtility.ConvertWindowsPath(dbsnpReferenceVcfPath) + "; fi",
// call variants
"if [ ! -f " + WrapperUtility.ConvertWindowsPath(HaplotypeCallerGvcfPath) + " ] || [ " + " ! -s " + WrapperUtility.ConvertWindowsPath(HaplotypeCallerGvcfPath) + " ]; then " +
Gatk(Workers, 2) +
" HaplotypeCaller" +
" --native-pair-hmm-threads " + threads.ToString() +
" -R " + WrapperUtility.ConvertWindowsPath(genomeFasta) +
" -I " + WrapperUtility.ConvertWindowsPath(splitTrimBam) +
" --min-base-quality-score 20" +
(experimentType == ExperimentType.RNASequencing ? " --dont-use-soft-clipped-bases true" : "") +
" --dbsnp " + WrapperUtility.ConvertWindowsPath(dbsnpReferenceVcfPath) +
" -O " + WrapperUtility.ConvertWindowsPath(HaplotypeCallerGvcfPath) +
" -ERC GVCF" + // this prompts phasing!
" --max-mnp-distance 3" + // note: this can't be used for joint genotyping here, but this setting is available in mutect2 for doing tumor vs normal calls
"; fi",
// index compressed gvcf file
$"if [ ! -f {WrapperUtility.ConvertWindowsPath($"{HaplotypeCallerGvcfPath}.tbi")} ]; then {Gatk(Workers)} IndexFeatureFile -F {WrapperUtility.ConvertWindowsPath(HaplotypeCallerGvcfPath)}; fi",
// genotype the gvcf file into a traditional vcf file
"if [ ! -f " + WrapperUtility.ConvertWindowsPath(HaplotypeCallerVcfPath) + " ] || [ " + " ! -s " + WrapperUtility.ConvertWindowsPath(HaplotypeCallerVcfPath) + " ]; then " +
Gatk(Workers, 2) +
" GenotypeGVCFs" +
" -R " + WrapperUtility.ConvertWindowsPath(genomeFasta) +
" -V " + WrapperUtility.ConvertWindowsPath(HaplotypeCallerGvcfPath) +
" -O " + WrapperUtility.ConvertWindowsPath(HaplotypeCallerVcfPath) +
"; fi",
$"if [ ! -f {WrapperUtility.ConvertWindowsPath($"{HaplotypeCallerVcfPath}.idx")} ]; then {Gatk(Workers)} IndexFeatureFile -F {WrapperUtility.ConvertWindowsPath(HaplotypeCallerVcfPath)}; fi",
// filter out indels
"if [ ! -f " + WrapperUtility.ConvertWindowsPath(FilteredHaplotypeCallerVcfPath) + " ] || [ " + " ! -s " + WrapperUtility.ConvertWindowsPath(FilteredHaplotypeCallerVcfPath) + " ]; then " +
Gatk(Workers, 2) +
" SelectVariants" +
" --select-type-to-exclude INDEL" +
" -R " + WrapperUtility.ConvertWindowsPath(genomeFasta) +
" -V " + WrapperUtility.ConvertWindowsPath(HaplotypeCallerVcfPath) +
" -O " + WrapperUtility.ConvertWindowsPath(FilteredHaplotypeCallerVcfPath) +
"; fi",
$"if [ ! -f {WrapperUtility.ConvertWindowsPath($"{FilteredHaplotypeCallerVcfPath}.idx")} ]; then {Gatk(Workers)} IndexFeatureFile -F {WrapperUtility.ConvertWindowsPath(FilteredHaplotypeCallerVcfPath)}; fi",
// filter variants (RNA-Seq specific params... need to check out recommendations before using DNA-Seq)
//"if [ ! -f " + WrapperUtility.ConvertWindowsPath(newVcf) + " ] || [ " + " ! -s " + WrapperUtility.ConvertWindowsPath(newVcf) + " ]; then " +
// Gatk() +
// " -T VariantFiltration" +
// " -nct " + threads.ToString() +
// " -R " + WrapperUtility.ConvertWindowsPath(genomeFasta) +
// " -V " + WrapperUtility.ConvertWindowsPath(unfliteredVcf) +
// " -window 35 -cluster 3" + // filter out clusters of 3 snps within 35 bases (https://software.broadinstitute.org/gatk/documentation/topic?name=methods)
// " -filterName FS -filter \"FS > 30.0\"" +
// " -filterName QD -filter \"QD < 2.0\"" +
// " -o " + WrapperUtility.ConvertWindowsPath(newVcf) +
// "; fi",
};
return(commands);
}
