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

Added Hard Clipping Tail Ends 

Added functionality to hard clip the low quality tail ends of reads (lowQual <= 2)
This commit is contained in:
Mauricio Carneiro 2011-08-15 15:22:54 -04:00
parent 0d976d6211
commit 993ecb85da
1 changed files with 27 additions and 2 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

29
public/java/src/org/broadinstitute/sting/utils/clipreads/ReadClipper.java
View File

@ -15,6 +15,7 @@ import java.util.List;
*/ */
public class ReadClipper { public class ReadClipper {
SAMRecord read; SAMRecord read;
boolean wasClipped;
List<ClippingOp> ops = null; List<ClippingOp> ops = null;
/** /**
@ -24,6 +25,7 @@ public class ReadClipper {
*/ */
public ReadClipper(final SAMRecord read) { public ReadClipper(final SAMRecord read) {
this.read = read; this.read = read;
this.wasClipped = false;
} }
/** /**
@ -41,7 +43,7 @@ public class ReadClipper {
} }
public boolean wasClipped() { public boolean wasClipped() {
return ops != null; return wasClipped;
} }
public SAMRecord getRead() { public SAMRecord getRead() {
@ -52,7 +54,7 @@ public class ReadClipper {
int start = (refStart < 0) ? 0 : ReadUtils.getReadCoordinateForReferenceCoordinate(read, refStart); int start = (refStart < 0) ? 0 : ReadUtils.getReadCoordinateForReferenceCoordinate(read, refStart);
int stop = (refStop < 0) ? read.getReadLength() - 1 : ReadUtils.getReadCoordinateForReferenceCoordinate(read, refStop); int stop = (refStop < 0) ? read.getReadLength() - 1 : ReadUtils.getReadCoordinateForReferenceCoordinate(read, refStop);
System.out.println("DEBUG -- clipping start/stop: " + start + "/" + stop); System.out.println("Clipping start/stop: " + start + "/" + stop);
this.addOp(new ClippingOp(start, stop)); this.addOp(new ClippingOp(start, stop));
return clipRead(ClippingRepresentation.HARDCLIP_BASES); return clipRead(ClippingRepresentation.HARDCLIP_BASES);
} }
@ -68,6 +70,28 @@ public class ReadClipper {
return hardClipByReferenceCoordinates(right, -1); return hardClipByReferenceCoordinates(right, -1);
} }
public SAMRecord hardClipLowQualEnds(byte lowQual) {
byte [] quals = read.getBaseQualities();
int leftClipIndex = 0;
int rightClipIndex = read.getReadLength() - 1;
// check how far we can clip both sides
while (quals[rightClipIndex] <= lowQual) rightClipIndex--;
while (quals[leftClipIndex] <= lowQual) leftClipIndex++;
// if the entire read should be clipped, then return an empty read. (--todo: maybe null is better? testing this for now)
if (leftClipIndex > rightClipIndex)
return (new SAMRecord(read.getHeader()));
if (rightClipIndex < read.getReadLength() - 1) {
this.addOp(new ClippingOp(rightClipIndex + 1, read.getReadLength() - 1));
}
if (leftClipIndex > 0 ) {
this.addOp(new ClippingOp(0, leftClipIndex - 1));
}
return this.clipRead(ClippingRepresentation.HARDCLIP_BASES);
}
/** /**
* Return a new read corresponding to this.read that's been clipped according to ops, if any are present. * Return a new read corresponding to this.read that's been clipped according to ops, if any are present.
* *
@ -83,6 +107,7 @@ public class ReadClipper {
for (ClippingOp op : getOps()) { for (ClippingOp op : getOps()) {
clippedRead = op.apply(algorithm, clippedRead); clippedRead = op.apply(algorithm, clippedRead);
} }
wasClipped = true;
return clippedRead; return clippedRead;
} catch (CloneNotSupportedException e) { } catch (CloneNotSupportedException e) {
throw new RuntimeException(e); // this should never happen throw new RuntimeException(e); // this should never happen