zzh
/
gatk-3.8
1
0
Fork 0
Code Issues Pull Requests Packages Projects Releases Wiki Activity

One more quick memory improvement: reuse Alleles in a given context instead of creating new ones for each sample (duh). 

git-svn-id: file:///humgen/gsa-scr1/gsa-engineering/svn_contents/trunk@3147 348d0f76-0448-11de-a6fe-93d51630548a
This commit is contained in:
ebanks 2010-04-09 18:48:36 +00:00
parent c2a37e4b5c
commit 0cc6d0fbbb
1 changed files with 25 additions and 18 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

43
java/src/org/broadinstitute/sting/gatk/refdata/VariantContextAdaptors.java
View File

@ -386,47 +386,54 @@ public class VariantContextAdaptors {
VariantContext convert(String name, Object input, Allele refAllele) { VariantContext convert(String name, Object input, Allele refAllele) {
PlinkRod plink = (PlinkRod)input; PlinkRod plink = (PlinkRod)input;
HashSet<Allele> alleles = new HashSet<Allele>(); HashSet<Allele> VCAlleles = new HashSet<Allele>();
alleles.add(refAllele); VCAlleles.add(refAllele);
// mapping from Plink Allele to VC Allele (since the PlinkRod does not annotate any of the Alleles as reference)
HashMap<Allele, Allele> plinkAlleleToVCAllele = new HashMap<Allele, Allele>();
Set<Genotype> genotypes = new HashSet<Genotype>(); Set<Genotype> genotypes = new HashSet<Genotype>();
Map<String, Allele[]> genotypeSets = plink.getGenotypes(); Map<String, Allele[]> genotypeSets = plink.getGenotypes();
// We need to iterate through this list and recreate the Alleles since the // We need to iterate through this list and recreate the Alleles since the
// PlinkRod does not promise to annotate any of the Alleles as reference // PlinkRod does not annotate any of the Alleles as reference.
// for each sample... // for each sample...
for ( Map.Entry<String, Allele[]> genotype : genotypeSets.entrySet() ) { for ( Map.Entry<String, Allele[]> genotype : genotypeSets.entrySet() ) {
ArrayList<Allele> myAlleles = new ArrayList<Allele>(2); ArrayList<Allele> myVCAlleles = new ArrayList<Allele>(2);
// for each allele... // for each allele...
for ( Allele myAllele : genotype.getValue() ) { for ( Allele myPlinkAllele : genotype.getValue() ) {
Allele allele; Allele VCAllele;
if ( myAllele.isNoCall() ) { if ( myPlinkAllele.isNoCall() ) {
allele = Allele.NO_CALL; VCAllele = Allele.NO_CALL;
} else { } else {
if ( !plink.isIndel() ) { if ( plinkAlleleToVCAllele.containsKey(myPlinkAllele) ) {
allele = new Allele(myAllele.getBases(), refAllele.equals(myAllele, true)); VCAllele = plinkAlleleToVCAllele.get(myPlinkAllele);
} else if ( myAllele.isNull() ) {
allele = new Allele(Allele.NULL_ALLELE_STRING, plink.isInsertion());
} else { } else {
allele = new Allele(myAllele.getBases(), !plink.isInsertion()); if ( !plink.isIndel() ) {
VCAllele = new Allele(myPlinkAllele.getBases(), refAllele.equals(myPlinkAllele, true));
} else if ( myPlinkAllele.isNull() ) {
VCAllele = new Allele(Allele.NULL_ALLELE_STRING, plink.isInsertion());
} else {
VCAllele = new Allele(myPlinkAllele.getBases(), !plink.isInsertion());
}
plinkAlleleToVCAllele.put(myPlinkAllele, VCAllele);
VCAlleles.add(VCAllele);
} }
alleles.add(allele);
} }
myAlleles.add(allele); myVCAlleles.add(VCAllele);
} }
// create the genotype // create the genotype
genotypes.add(new Genotype(genotype.getKey(), myAlleles)); genotypes.add(new Genotype(genotype.getKey(), myVCAlleles));
} }
// create the variant context // create the variant context
try { try {
GenomeLoc loc = GenomeLocParser.setStop(plink.getLocation(), plink.getLocation().getStop() + plink.getLength()-1); GenomeLoc loc = GenomeLocParser.setStop(plink.getLocation(), plink.getLocation().getStop() + plink.getLength()-1);
VariantContext vc = new VariantContext(plink.getVariantName(), loc, alleles, genotypes); VariantContext vc = new VariantContext(plink.getVariantName(), loc, VCAlleles, genotypes);
vc.validate(); vc.validate();
return vc; return vc;
} catch (IllegalArgumentException e) { } catch (IllegalArgumentException e) {