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

Adding new command line argument requested on the forum to control the maximum number of haplotypes that are sent forward for genotyping. In the presence of a large degree of heterozygosity the current algorithm breaks down and so this argument would need to be increased.

This commit is contained in:
Ryan Poplin 2013-01-09 15:54:02 -05:00
parent 487fb2afb4
commit 1a18947abf
2 changed files with 8 additions and 3 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

7
protected/java/src/org/broadinstitute/sting/gatk/walkers/haplotypecaller/HaplotypeCaller.java
View File

@ -136,6 +136,10 @@ public class HaplotypeCaller extends ActiveRegionWalker<Integer, Integer> implem
@Argument(fullName="gcpHMM", shortName="gcpHMM", doc="Flat gap continuation penalty for use in the Pair HMM", required = false)
protected int gcpHMM = 10;
@Advanced
@Argument(fullName="maxNumHaplotypesInPopulation", shortName="maxNumHaplotypesInPopulation", doc="Maximum number of haplotypes to consider for your population. This number will probably need to be increased when calling organisms with high heterozygosity.", required = false)
protected int maxNumHaplotypesInPopulation = 13;
@Advanced
@Argument(fullName="minKmer", shortName="minKmer", doc="Minimum kmer length to use in the assembly graph", required = false)
protected int minKmer = 11;
@ -414,7 +418,8 @@ public class HaplotypeCaller extends ActiveRegionWalker<Integer, Integer> implem
final Map<String, ArrayList<GATKSAMRecord>> perSampleFilteredReadList = splitReadsBySample( filteredReads );
// subset down to only the best haplotypes to be genotyped in all samples ( in GGA mode use all discovered haplotypes )
final ArrayList<Haplotype> bestHaplotypes = ( UG_engine.getUAC().GenotypingMode != GenotypeLikelihoodsCalculationModel.GENOTYPING_MODE.GENOTYPE_GIVEN_ALLELES ? likelihoodCalculationEngine.selectBestHaplotypes( haplotypes, stratifiedReadMap ) : haplotypes );
final ArrayList<Haplotype> bestHaplotypes = ( UG_engine.getUAC().GenotypingMode != GenotypeLikelihoodsCalculationModel.GENOTYPING_MODE.GENOTYPE_GIVEN_ALLELES ?
likelihoodCalculationEngine.selectBestHaplotypes( haplotypes, stratifiedReadMap, maxNumHaplotypesInPopulation ) : haplotypes );
for( final VariantContext call : genotypingEngine.assignGenotypeLikelihoods( UG_engine,
bestHaplotypes,

4
protected/java/src/org/broadinstitute/sting/gatk/walkers/haplotypecaller/LikelihoodCalculationEngine.java
View File

@ -273,7 +273,7 @@ public class LikelihoodCalculationEngine {
@Requires({"haplotypes.size() > 0"})
@Ensures({"result.size() <= haplotypes.size()"})
public ArrayList<Haplotype> selectBestHaplotypes( final ArrayList<Haplotype> haplotypes, final Map<String, PerReadAlleleLikelihoodMap> stratifiedReadMap ) {
public ArrayList<Haplotype> selectBestHaplotypes( final ArrayList<Haplotype> haplotypes, final Map<String, PerReadAlleleLikelihoodMap> stratifiedReadMap, final int maxNumHaplotypesInPopulation ) {
final int numHaplotypes = haplotypes.size();
final Set<String> sampleKeySet = stratifiedReadMap.keySet();
@ -287,7 +287,7 @@ public class LikelihoodCalculationEngine {
int hap1 = 0;
int hap2 = 0;
//double bestElement = Double.NEGATIVE_INFINITY;
final int maxChosenHaplotypes = Math.min( 13, sampleKeySet.size() * 2 + 1 );
final int maxChosenHaplotypes = Math.min( maxNumHaplotypesInPopulation, sampleKeySet.size() * 2 + 1 );
while( bestHaplotypesIndexList.size() < maxChosenHaplotypes ) {
double maxElement = Double.NEGATIVE_INFINITY;
for( int iii = 0; iii < numHaplotypes; iii++ ) {