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

added naive ad-hoc cutoff for the pile size the cleaner will attempt to process; use --maxPileSize argument to force any pile larger than specified cutoff to be directly written to the output without cleaning 

git-svn-id: file:///humgen/gsa-scr1/gsa-engineering/svn_contents/trunk@972 348d0f76-0448-11de-a6fe-93d51630548a
This commit is contained in:
asivache 2009-06-10 17:52:35 +00:00
parent f9be175f44
commit 030efc468f
1 changed files with 26 additions and 6 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

28
java/src/org/broadinstitute/sting/playground/gatk/walkers/indels/IntervalCleanerWalker.java
View File

@ -11,10 +11,7 @@ import org.broadinstitute.sting.playground.indels.*;
import net.sf.samtools.*; import net.sf.samtools.*;
import java.util.*; import java.util.*;
import java.io.File; import java.io.File;
import java.io.FileNotFoundException;
import java.io.FileOutputStream;
import java.io.FileWriter; import java.io.FileWriter;
import java.io.OutputStream;
@WalkerName("IntervalCleaner") @WalkerName("IntervalCleaner")
public class IntervalCleanerWalker extends LocusWindowWalker<Integer, Integer> { public class IntervalCleanerWalker extends LocusWindowWalker<Integer, Integer> {
@ -30,6 +27,8 @@ public class IntervalCleanerWalker extends LocusWindowWalker<Integer, Integer>
public String OUT_STATS = null; public String OUT_STATS = null;
@Argument(fullName="LODThresholdForCleaning", shortName="LOD", doc="LOD threshold above which the cleaner will clean", required=false) @Argument(fullName="LODThresholdForCleaning", shortName="LOD", doc="LOD threshold above which the cleaner will clean", required=false)
public double LOD_THRESHOLD = 5.0; public double LOD_THRESHOLD = 5.0;
@Argument(fullName="maxPileSize", shortName="maxSize", doc="max number of reads in the pile; if exceeded, no attempt will be made to realign the pile", required=false)
public int maxPileSize = 1000000000;
@Argument(fullName="EntropyThreshold", shortName="entropy", doc="percentage of mismatches at a locus to be considered having high entropy", required=false) @Argument(fullName="EntropyThreshold", shortName="entropy", doc="percentage of mismatches at a locus to be considered having high entropy", required=false)
public double MISMATCH_THRESHOLD = 0.25; public double MISMATCH_THRESHOLD = 0.25;
@Argument(fullName="GreedyThreshold", shortName="greedy", doc="coverage above which the cleaner turns on greedy mode to improve performance", required=false) @Argument(fullName="GreedyThreshold", shortName="greedy", doc="coverage above which the cleaner turns on greedy mode to improve performance", required=false)
@ -109,7 +108,28 @@ public class IntervalCleanerWalker extends LocusWindowWalker<Integer, Integer>
else else
readsToWrite.add(new ComparableSAMRecord(read)); readsToWrite.add(new ComparableSAMRecord(read));
} }
clean(goodReads, ref, context.getLocation());
if ( goodReads.size() > maxPileSize ) {
// too many reads, shy away!
if ( statsOutput != null ) {
try {
statsOutput.write(context.getLocation().toString());
statsOutput.write("\tSKIPPED ("+reads.size()+" reads total, "+goodReads.size()+" for realignment)\t");
statsOutput.write("-1.0");
statsOutput.write("\n");
statsOutput.flush();
} catch (Exception e) {}
}
// push all "good" reads into readsToWrite without cleaning, there are too many!
for ( SAMRecord read : goodReads ) {
readsToWrite.add(new ComparableSAMRecord(read));
}
goodReads.clear();
} else {
clean(goodReads, ref, context.getLocation());
}
//bruteForceClean(goodReads, ref, context.getLocation().getStart()); //bruteForceClean(goodReads, ref, context.getLocation().getStart());
//testCleanWithDeletion(); //testCleanWithDeletion();
//testCleanWithInsertion(); //testCleanWithInsertion();