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

When crossing contigs it's crucial that the queue get flushed or else it will continue to accumulate reads without emitting. This is the last time I trust someone when they tell me that they are 'confident there are no bugs' in a tool. 

git-svn-id: file:///humgen/gsa-scr1/gsa-engineering/svn_contents/trunk@5315 348d0f76-0448-11de-a6fe-93d51630548a
This commit is contained in:
ebanks 2011-02-25 05:18:30 +00:00
parent 44a48b4178
commit 5d28cbda27
1 changed files with 26 additions and 29 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

55
java/src/org/broadinstitute/sting/utils/sam/ConstrainedMateFixingSAMFileWriter.java
View File

@ -4,6 +4,7 @@ import net.sf.picard.sam.SamPairUtil;
import net.sf.samtools.*; import net.sf.samtools.*;
import org.apache.log4j.Logger; import org.apache.log4j.Logger;
import org.broadinstitute.sting.utils.exceptions.ReviewedStingException; import org.broadinstitute.sting.utils.exceptions.ReviewedStingException;
//import org.broadinstitute.sting.utils.SimpleTimer;
import java.io.File; import java.io.File;
import java.util.*; import java.util.*;
@ -104,9 +105,12 @@ public class ConstrainedMateFixingSAMFileWriter implements SAMFileWriter {
/** read.name -> records */ /** read.name -> records */
HashMap<String, SAMRecord> forMateMatching = new HashMap<String, SAMRecord>(); HashMap<String, SAMRecord> forMateMatching = new HashMap<String, SAMRecord>();
//Queue<SAMRecord> waitingReads = new LinkedList<SAMRecord>();
Queue<SAMRecord> waitingReads = new PriorityQueue<SAMRecord>(1000, comparer); Queue<SAMRecord> waitingReads = new PriorityQueue<SAMRecord>(1000, comparer);
//private SimpleTimer timer = new SimpleTimer("ConstrainedWriter");
//private long PROGRESS_PRINT_FREQUENCY = 10 * 1000; // in milliseconds
//private long lastProgressPrintTime = -1; // When was the last time we printed progress log?
/** /**
* *
@ -127,6 +131,9 @@ public class ConstrainedMateFixingSAMFileWriter implements SAMFileWriter {
final int maxInsertSizeForMovingReadPairs) { final int maxInsertSizeForMovingReadPairs) {
this.finalDestination = finalDestination; this.finalDestination = finalDestination;
this.maxInsertSizeForMovingReadPairs = maxInsertSizeForMovingReadPairs; this.maxInsertSizeForMovingReadPairs = maxInsertSizeForMovingReadPairs;
//timer.start();
//lastProgressPrintTime = timer.currentTime();
} }
public int getMaxReadsInQueue() { return maxReadsInQueue; } public int getMaxReadsInQueue() { return maxReadsInQueue; }
@ -144,33 +151,6 @@ public class ConstrainedMateFixingSAMFileWriter implements SAMFileWriter {
return pos + 2 * MAX_POS_MOVE_ALLOWED < addedRead.getAlignmentStart(); return pos + 2 * MAX_POS_MOVE_ALLOWED < addedRead.getAlignmentStart();
} }
// private void verifyOrdering() {
// SAMRecord lastRead = null;
// List<SAMRecord> reads = new ArrayList<SAMRecord>();
//
// reads.addAll(waitingReads);
// Collections.sort(reads, comparer);
// for ( SAMRecord read : reads ) {
// logger.info("READ is " + read.getReadName() + " pos " + read.getAlignmentStart());
// if ( lastRead != null && comparer.fileOrderCompare(lastRead, read) > 0 )
// throw new ReviewedStingException("BUG: records added out of order: read1=" + lastRead +
// ", pos=" + lastRead.getAlignmentStart() + " read2="+read + ", pos=" + read.getAlignmentStart());
// lastRead = read;
// }
//
//// List<SAMRecord> reads = new ArrayList<SAMRecord>();
//// while ( waitingReads.peek() != null ) {
//// SAMRecord read = waitingReads.poll();
//// logger.info("READ is " + read.getReadName() + " pos " + read.getAlignmentStart());
//// if ( lastRead != null && comparer.fileOrderCompare(lastRead, read) > 0 )
//// throw new ReviewedStingException("BUG: records added out of order: read1=" + lastRead +
//// ", pos=" + lastRead.getAlignmentStart() + " read2="+read + ", pos=" + read.getAlignmentStart());
//// lastRead = read;
//// reads.add(read);
//// }
//
// for ( SAMRecord read : reads ) waitingReads.add(read);
// }
/** /**
* @{inheritDoc} * @{inheritDoc}
@ -178,6 +158,24 @@ public class ConstrainedMateFixingSAMFileWriter implements SAMFileWriter {
public void addAlignment( SAMRecord newRead ) { public void addAlignment( SAMRecord newRead ) {
if ( DEBUG ) logger.info("New read pos " + newRead.getAlignmentStart()); if ( DEBUG ) logger.info("New read pos " + newRead.getAlignmentStart());
//final long curTime = timer.currentTime();
//if ( curTime - lastProgressPrintTime > PROGRESS_PRINT_FREQUENCY ) {
// lastProgressPrintTime = curTime;
// System.out.println("WaitingReads.size = " + waitingReads.size() + ", forMateMatching.size = " + forMateMatching.size());
//}
// if the new read is on a different contig, then we need to flush the queue and clear the map
if ( waitingReads.size() > 0 && waitingReads.peek().getReferenceIndex() != newRead.getReferenceIndex()) {
if ( DEBUG ) logger.warn("Flushing queue on move to new contig: " + newRead.getReferenceName());
while ( ! waitingReads.isEmpty() ) {
// emit to disk
finalDestination.addAlignment(waitingReads.remove());
}
forMateMatching.clear();
}
// fix mates, as needed // fix mates, as needed
// Since setMateInfo can move reads, we potentially need to remove the mate, and requeue // Since setMateInfo can move reads, we potentially need to remove the mate, and requeue
// it to ensure proper sorting // it to ensure proper sorting
@ -206,7 +204,6 @@ public class ConstrainedMateFixingSAMFileWriter implements SAMFileWriter {
maxReadsInQueue = Math.max(maxReadsInQueue, waitingReads.size()); maxReadsInQueue = Math.max(maxReadsInQueue, waitingReads.size());
if ( ++counter % EMIT_FREQUENCY == 0 ) { if ( ++counter % EMIT_FREQUENCY == 0 ) {
//verifyOrdering();
while ( ! waitingReads.isEmpty() ) { // there's something in the queue while ( ! waitingReads.isEmpty() ) { // there's something in the queue
SAMRecord read = waitingReads.peek(); SAMRecord read = waitingReads.peek();