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

New mode for CombineVariants to assume the incoming VCFs have the same samples and disjoint calls. Drastically reduces the runtime for routine combining operations. Very useful with Queue. 

git-svn-id: file:///humgen/gsa-scr1/gsa-engineering/svn_contents/trunk@5356 348d0f76-0448-11de-a6fe-93d51630548a
This commit is contained in:
rpoplin 2011-03-02 15:52:17 +00:00
parent 5e4b321f86
commit 8e1aa6059a
1 changed files with 12 additions and 0 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

12
java/src/org/broadinstitute/sting/gatk/walkers/variantutils/CombineVariants.java
View File

@ -79,6 +79,9 @@ public class CombineVariants extends RodWalker<Integer, Integer> {
@Argument(fullName="setKey", shortName="setKey", doc="Key, by default set, in the INFO key=value tag emitted describing which set the combined VCF record came from. Set to null if you don't want the set field emitted.", required=false)
public String SET_KEY = "set";
@Argument(fullName="assumeIdenticalSamples", shortName="assumeIdenticalSamples", doc="If true, assume input VCFs have identical sample sets and disjoint calls so that one can simply perform a merge sort to combine the VCFs into one, drastically reducing the runtime.", required=false)
public boolean ASSUME_IDENTICAL_SAMPLES = false;
private List<String> priority = null;
public void initialize() {
@ -127,6 +130,15 @@ public class CombineVariants extends RodWalker<Integer, Integer> {
// Need to provide reference bases to simpleMerge starting at current locus
Collection<VariantContext> vcs = tracker.getAllVariantContexts(ref, context.getLocation());
if ( ASSUME_IDENTICAL_SAMPLES ) {
final VariantContext vc = vcs.iterator().next();
if( vc != null ) {
vcfWriter.add( vc, ref.getBase() );
}
return vcs.isEmpty() ? 0 : 1;
}
VariantContext mergedVC = null;
if ( variantMergeOption == VariantContextUtils.VariantMergeType.MASTER ) {
mergedVC = VariantContextUtils.masterMerge(vcs, "master");