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

Cleanup. Move filtering functions closer to where they're used. 

git-svn-id: file:///humgen/gsa-scr1/gsa-engineering/svn_contents/trunk@767 348d0f76-0448-11de-a6fe-93d51630548a
This commit is contained in:
hanna 2009-05-20 21:42:48 +00:00
parent e7a6f8cdc4
commit 26dd3cd50e
2 changed files with 42 additions and 49 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

44
java/src/org/broadinstitute/sting/gatk/dataSources/providers/LocusContextQueue.java
View File

@ -5,13 +5,14 @@ import org.broadinstitute.sting.gatk.Reads;
import org.broadinstitute.sting.gatk.dataSources.shards.Shard;
import org.broadinstitute.sting.gatk.iterators.LocusContextIteratorByHanger;
import org.broadinstitute.sting.gatk.iterators.LocusContextIterator;
import org.broadinstitute.sting.gatk.traversals.TraversalEngine;
import org.broadinstitute.sting.gatk.traversals.TraversalStatistics;
import org.broadinstitute.sting.utils.GenomeLoc;
import net.sf.samtools.SAMRecord;
import java.util.Iterator;
import edu.mit.broad.picard.filter.FilteringIterator;
import edu.mit.broad.picard.filter.SamRecordFilter;
/**
* User: hanna
* Date: May 13, 2009
@ -36,7 +37,7 @@ public abstract class LocusContextQueue {
private LocusContextIterator loci;
public LocusContextQueue(ShardDataProvider provider) {
Iterator<SAMRecord> reads = new FilteringIterator(provider.getReadIterator(), new TraversalEngine.locusStreamFilterFunc());
Iterator<SAMRecord> reads = new FilteringIterator(provider.getReadIterator(), new LocusStreamFilterFunc());
this.loci = new LocusContextIteratorByHanger(reads);
this.sourceInfo = provider.getReadIterator().getSourceInfo();
this.shard = provider.getShard();
@ -75,4 +76,43 @@ public abstract class LocusContextQueue {
return loci.hasNext();
}
/**
* Class to filter out un-handle-able reads from the stream. We currently are skipping
* unmapped reads, non-primary reads, unaligned reads, and duplicate reads.
*/
private static class LocusStreamFilterFunc implements SamRecordFilter {
SAMRecord lastRead = null;
public boolean filterOut(SAMRecord rec) {
boolean result = false;
String why = "";
if (rec.getReadUnmappedFlag()) {
TraversalStatistics.nUnmappedReads++;
result = true;
why = "Unmapped";
} else if (rec.getNotPrimaryAlignmentFlag()) {
TraversalStatistics.nNotPrimary++;
result = true;
why = "Not Primary";
} else if (rec.getAlignmentStart() == SAMRecord.NO_ALIGNMENT_START) {
TraversalStatistics.nBadAlignments++;
result = true;
why = "No alignment start";
} else if (rec.getDuplicateReadFlag()) {
TraversalStatistics.nDuplicates++;
result = true;
why = "Duplicate reads";
}
else {
result = false;
}
if (result) {
TraversalStatistics.nSkippedReads++;
//System.out.printf(" [filter] %s => %b %s", rec.getReadName(), result, why);
} else {
TraversalStatistics.nReads++;
}
return result;
}
}
}

47
java/src/org/broadinstitute/sting/gatk/traversals/TraversalEngine.java
View File

@ -546,53 +546,6 @@ public abstract class TraversalEngine {
throw new UnsupportedOperationException( "This method is a required override for new traversal engines. Please port your traversal engine to the new style." );
}
// --------------------------------------------------------------------------------------------------------------
//
// traversal by loci functions
//
// --------------------------------------------------------------------------------------------------------------
/**
* Class to filter out un-handle-able reads from the stream. We currently are skipping
* unmapped reads, non-primary reads, unaligned reads, and duplicate reads.
*/
public static class locusStreamFilterFunc implements SamRecordFilter {
SAMRecord lastRead = null;
public boolean filterOut(SAMRecord rec) {
boolean result = false;
String why = "";
if (rec.getReadUnmappedFlag()) {
TraversalStatistics.nUnmappedReads++;
result = true;
why = "Unmapped";
} else if (rec.getNotPrimaryAlignmentFlag()) {
TraversalStatistics.nNotPrimary++;
result = true;
why = "Not Primary";
} else if (rec.getAlignmentStart() == SAMRecord.NO_ALIGNMENT_START) {
TraversalStatistics.nBadAlignments++;
result = true;
why = "No alignment start";
} else if (rec.getDuplicateReadFlag()) {
TraversalStatistics.nDuplicates++;
result = true;
why = "Duplicate reads";
}
else {
result = false;
}
if (result) {
TraversalStatistics.nSkippedReads++;
//System.out.printf(" [filter] %s => %b %s", rec.getReadName(), result, why);
} else {
TraversalStatistics.nReads++;
}
return result;
}
}
public void verifySortOrder(final boolean requiresSortedOrder) {
if (beSafeP && !SORT_ON_FLY && samReader.getFileHeader().getSortOrder() != SAMFileHeader.SortOrder.coordinate) {
final String msg = "SAM file is not sorted in coordinate order (according to header) Walker type with given arguments requires a sorted file for correct processing";