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

Revert "ReduceReads ref bases are now output as '=' 

Reducing the reference bases to '=' results in an extra compression of 13% on average. The GATK is not ready to handle files with '=' bases, and the decision was to implement this a an engine support, not a part of ReduceReads.
This commit is contained in:
Mauricio Carneiro 2011-11-01 16:35:07 -04:00
parent 76c32f5409
commit 17cc484dbd
1 changed files with 745 additions and 756 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

23
public/java/src/org/broadinstitute/sting/utils/sam/AlignmentUtils.java
View File

@ -198,13 +198,15 @@ public class AlignmentUtils {
return sum; return sum;
} }
/** Returns the mismatches in the read within the given reference context. /** Returns the number of mismatches in the pileup element within the given reference context.
* *
* @param read the SAMRecord * @param read the SAMRecord
* @param ref the reference context * @param ref the reference context
* @return each base is represented by a bit in the BitSet. True for mismatch, false for ref * @param maxMismatches the maximum number of surrounding mismatches we tolerate to consider a base good
* @param windowSize window size (on each side) to test
* @return a bitset representing which bases are good
*/ */
public static BitSet mismatchesInRefWindow(SAMRecord read, ReferenceContext ref) { public static BitSet mismatchesInRefWindow(SAMRecord read, ReferenceContext ref, int maxMismatches, int windowSize) {
// first determine the positions with mismatches // first determine the positions with mismatches
int readLength = read.getReadLength(); int readLength = read.getReadLength();
BitSet mismatches = new BitSet(readLength); BitSet mismatches = new BitSet(readLength);
@ -212,7 +214,7 @@ public class AlignmentUtils {
// it's possible we aren't starting at the beginning of a read, // it's possible we aren't starting at the beginning of a read,
// and we don't need to look at any of the previous context outside our window // and we don't need to look at any of the previous context outside our window
// (although we do need future context) // (although we do need future context)
int readStartPos = read.getAlignmentStart(); int readStartPos = Math.max(read.getAlignmentStart(), ref.getLocus().getStart() - windowSize);
int currentReadPos = read.getAlignmentStart(); int currentReadPos = read.getAlignmentStart();
byte[] refBases = ref.getBases(); byte[] refBases = ref.getBases();
@ -263,22 +265,9 @@ public class AlignmentUtils {
break; break;
} }
} }
return mismatches;
}
/** Returns a BitSet showing what bases are good according to the criteria of maxMismatches in the window
*
* @param read the SAMRecord
* @param ref the reference context
* @param maxMismatches the maximum number of surrounding mismatches we tolerate to consider a base good
* @param windowSize window size (on each side) to test
* @return a bitset representing which bases are good
*/
public static BitSet mismatchesInRefWindow(SAMRecord read, ReferenceContext ref, int maxMismatches, int windowSize) {
int readLength = read.getReadLength();
// all bits are set to false by default // all bits are set to false by default
BitSet result = new BitSet(readLength); BitSet result = new BitSet(readLength);
BitSet mismatches = mismatchesInRefWindow(read, ref);
int currentPos = 0, leftPos = 0, rightPos; int currentPos = 0, leftPos = 0, rightPos;
int mismatchCount = 0; int mismatchCount = 0;