From f448a263e982e30d85dbc1bf180e58728b9949aa Mon Sep 17 00:00:00 2001
From: ebanks <ebanks@348d0f76-0448-11de-a6fe-93d51630548a>
Date: Mon, 21 Dec 2009 21:01:55 +0000
Subject: [PATCH] The cleaner now cleans duplicate reads (instead of ignoring
 them) - although it doesn't include them for scoring ref or alt consenses

git-svn-id: file:///humgen/gsa-scr1/gsa-engineering/svn_contents/trunk@2424 348d0f76-0448-11de-a6fe-93d51630548a
---
 .../gatk/walkers/indels/IntervalCleanerWalker.java   | 12 +++++-------
 1 file changed, 5 insertions(+), 7 deletions(-)

diff --git a/java/src/org/broadinstitute/sting/gatk/walkers/indels/IntervalCleanerWalker.java b/java/src/org/broadinstitute/sting/gatk/walkers/indels/IntervalCleanerWalker.java
index 6006e664b..0897fead3 100755
--- a/java/src/org/broadinstitute/sting/gatk/walkers/indels/IntervalCleanerWalker.java
+++ b/java/src/org/broadinstitute/sting/gatk/walkers/indels/IntervalCleanerWalker.java
@@ -111,8 +111,7 @@ public class  IntervalCleanerWalker extends LocusWindowWalker<Integer, Integer>
         for ( SAMRecord read : reads ) {
             if ( !read.getReadUnmappedFlag() &&
                  !read.getNotPrimaryAlignmentFlag() &&
-                 !read.getDuplicateReadFlag() &&
-                 read.getMappingQuality() != 0 &&           
+                 read.getMappingQuality() != 0 &&
                  read.getAlignmentStart() != SAMRecord.NO_ALIGNMENT_START &&
 		         (allow454 || !Utils.is454Read(read)) )
                 goodReads.add(read);
@@ -121,9 +120,6 @@ public class  IntervalCleanerWalker extends LocusWindowWalker<Integer, Integer>
         }
 
         clean(goodReads, ref, loc);
-        //bruteForceClean(goodReads, ref, context.getLocation().getStart());
-        //testCleanWithDeletion();
-        //testCleanWithInsertion();
 
         if ( writer != null ) {
             // Although we can guarantee that reads will be emitted in order WITHIN an interval
@@ -257,7 +253,8 @@ public class  IntervalCleanerWalker extends LocusWindowWalker<Integer, Integer>
             // if this doesn't match perfectly to the reference, let's try to clean it
             if ( mismatchScore > 0 ) {
                 altReads.add(aRead);
-                totalMismatchSum += mismatchScore;
+                if ( !read.getDuplicateReadFlag() )
+                    totalMismatchSum += mismatchScore;
                 aRead.setMismatchScoreToReference(mismatchScore);
                 // if it has an indel, let's see if that's the best consensus
                 if ( numBlocks == 2 )
@@ -325,7 +322,8 @@ public class  IntervalCleanerWalker extends LocusWindowWalker<Integer, Integer>
                     consensus.readIndexes.add(new Pair<Integer, Integer>(j, altAlignment.first));
 
                 //logger.debug(consensus.str +  " vs. " + toTest.getRead().getReadString() + " => " + myScore + " - " + altAlignment.first);
-                consensus.mismatchSum += myScore;
+                if ( !toTest.getRead().getDuplicateReadFlag() )
+                    consensus.mismatchSum += myScore;
             }
 
             //logger.debug(consensus.str +  " " + consensus.mismatchSum);