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

Can now print out list of intervals and their stats 

git-svn-id: file:///humgen/gsa-scr1/gsa-engineering/svn_contents/trunk@5071 348d0f76-0448-11de-a6fe-93d51630548a
This commit is contained in:
fromer 2011-01-25 18:36:59 +00:00
parent 32cdcc933c
commit 17ba75e502
1 changed files with 28 additions and 10 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

38
java/src/org/broadinstitute/sting/oneoffprojects/phasing/CountHetPhasingInIntervalWalker.java
View File

@ -26,6 +26,7 @@ package org.broadinstitute.sting.oneoffprojects.phasing;
import org.broad.tribble.util.variantcontext.Genotype; import org.broad.tribble.util.variantcontext.Genotype;
import org.broad.tribble.util.variantcontext.VariantContext; import org.broad.tribble.util.variantcontext.VariantContext;
import org.broadinstitute.sting.commandline.Argument;
import org.broadinstitute.sting.commandline.Output; import org.broadinstitute.sting.commandline.Output;
import org.broadinstitute.sting.gatk.contexts.AlignmentContext; import org.broadinstitute.sting.gatk.contexts.AlignmentContext;
import org.broadinstitute.sting.gatk.contexts.ReferenceContext; import org.broadinstitute.sting.gatk.contexts.ReferenceContext;
@ -59,11 +60,14 @@ public class CountHetPhasingInIntervalWalker extends RodWalker<Integer, Integer>
@Output @Output
protected PrintStream out; protected PrintStream out;
@Argument(fullName = "perIntervalOut", shortName = "perIntervalOut", doc = "File to which to write per-sample, per-interval phased het statistics", required = false)
protected PrintStream perIntervalOut = null;
public void initialize() { public void initialize() {
rodNames = new LinkedList<String>(); rodNames = new LinkedList<String>();
rodNames.add("variant"); rodNames.add("variant");
intervalStats = new MultiSampleIntervalStats(); intervalStats = new MultiSampleIntervalStats(perIntervalOut);
} }
public boolean generateExtendedEvents() { public boolean generateExtendedEvents() {
@ -100,7 +104,7 @@ public class CountHetPhasingInIntervalWalker extends RodWalker<Integer, Integer>
boolean isNewInterval = (prevInterval == null || !curInterval.equals(prevInterval)); boolean isNewInterval = (prevInterval == null || !curInterval.equals(prevInterval));
if (isNewInterval) if (isNewInterval)
intervalStats.startNewInterval(); intervalStats.startNewInterval(curInterval);
boolean requireStartHere = true; // only see each VariantContext once boolean requireStartHere = true; // only see each VariantContext once
boolean takeFirstOnly = false; // take as many entries as the VCF file has boolean takeFirstOnly = false; // take as many entries as the VCF file has
@ -142,9 +146,13 @@ public class CountHetPhasingInIntervalWalker extends RodWalker<Integer, Integer>
private Map<String, SingleSampleIntervalStats> sampleToStat; private Map<String, SingleSampleIntervalStats> sampleToStat;
protected int numIntervals; protected int numIntervals;
public MultiSampleIntervalStats() { private PrintStream perIntervalOut;
private GenomeLoc curInterval;
public MultiSampleIntervalStats(PrintStream perIntervalOut) {
this.sampleToStat = new HashMap<String, SingleSampleIntervalStats>(); this.sampleToStat = new HashMap<String, SingleSampleIntervalStats>();
this.numIntervals = 0; this.numIntervals = 0;
this.perIntervalOut = perIntervalOut;
} }
public void processHetSiteInInterval(String sample, boolean isHet, boolean isPhased) { public void processHetSiteInInterval(String sample, boolean isHet, boolean isPhased) {
@ -158,19 +166,29 @@ public class CountHetPhasingInIntervalWalker extends RodWalker<Integer, Integer>
} }
public void finalizeStats() { public void finalizeStats() {
for (SingleSampleIntervalStats stats : sampleToStat.values()) if (curInterval == null)
stats.finalizeStats(); return;
for (Map.Entry<String, SingleSampleIntervalStats> sampleStatEntry : sampleToStat.entrySet()) {
SingleSampleIntervalStats stats = sampleStatEntry.getValue();
if (perIntervalOut != null) {
String sample = sampleStatEntry.getKey();
perIntervalOut.print(sample + "\t" + curInterval + "\t" + stats.numPhasedInCurrentInterval + "\t" + stats.numHetsInCurrentInterval + "\t" + stats.firstHetIsPhased);
}
stats.finalizeStats(); // now, can reset the counters [after print-out]
}
}
public void startNewInterval(GenomeLoc curInterval) {
finalizeStats();
numIntervals++;
this.curInterval = curInterval;
} }
public Set<Map.Entry<String, SingleSampleIntervalStats>> entrySet() { public Set<Map.Entry<String, SingleSampleIntervalStats>> entrySet() {
return sampleToStat.entrySet(); return sampleToStat.entrySet();
} }
public void startNewInterval() {
finalizeStats();
numIntervals++;
}
private class SingleSampleIntervalStats { private class SingleSampleIntervalStats {
public Map<PhasedHetsStat, Integer> hetStatInIntervalToCount; public Map<PhasedHetsStat, Integer> hetStatInIntervalToCount;
public int firstHetIsPhased; public int firstHetIsPhased;