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

Reduce unnecessary repetitive accesses to the BAM index file. 

git-svn-id: file:///humgen/gsa-scr1/gsa-engineering/svn_contents/trunk@5607 348d0f76-0448-11de-a6fe-93d51630548a
This commit is contained in:
hanna 2011-04-08 19:28:14 +00:00
parent 0a58d7aa1a
commit b4b52cc0fe
2 changed files with 23 additions and 8 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

3
java/src/org/broadinstitute/sting/gatk/datasources/reads/BAMSchedule.java
View File

@ -103,6 +103,7 @@ public class BAMSchedule implements CloseableIterator<BAMScheduleEntry> {
for(final SAMReaderID reader: readerIDs) { for(final SAMReaderID reader: readerIDs) {
GATKBAMIndex index = indices.get(reader); GATKBAMIndex index = indices.get(reader);
BAMIndexContent indexContent = index.getQueryResults(referenceSequence);
int currentBinInLowestLevel = GATKBAMIndex.getFirstBinInLevel(GATKBAMIndex.getNumIndexLevels()-1); int currentBinInLowestLevel = GATKBAMIndex.getFirstBinInLevel(GATKBAMIndex.getNumIndexLevels()-1);
Iterator<GenomeLoc> locusIterator = intervals.iterator(); Iterator<GenomeLoc> locusIterator = intervals.iterator();
@ -132,7 +133,7 @@ public class BAMSchedule implements CloseableIterator<BAMScheduleEntry> {
// Code at this point knows that the current bin is neither before nor after the current locus, // Code at this point knows that the current bin is neither before nor after the current locus,
// so it must overlap. Add this region to the filesystem. // so it must overlap. Add this region to the filesystem.
final GATKBAMFileSpan fileSpan = index.getSpanOverlapping(bin); final GATKBAMFileSpan fileSpan = index.getSpanOverlapping(indexContent,bin);
if(!fileSpan.isEmpty()) { if(!fileSpan.isEmpty()) {
// File format is binary in little endian; start of region, end of region, num chunks, then the chunks themselves. // File format is binary in little endian; start of region, end of region, num chunks, then the chunks themselves.

28
java/src/org/broadinstitute/sting/gatk/datasources/reads/GATKBAMIndex.java
View File

@ -304,7 +304,21 @@ public class GATKBAMIndex implements BAMIndex, BrowseableBAMIndex {
final int referenceSequence = gatkBin.getReferenceSequence(); final int referenceSequence = gatkBin.getReferenceSequence();
BAMIndexContent indexQuery = getQueryResults(referenceSequence); BAMIndexContent indexQuery = getQueryResults(referenceSequence);
if(indexQuery == null) return getSpanOverlapping(indexQuery,bin);
}
/**
* Perform an overlapping query of all bins bounding the given location.
* @param bin The bin over which to perform an overlapping query.
* @return The file pointers
*/
public GATKBAMFileSpan getSpanOverlapping(final BAMIndexContent indexContent, final Bin bin) {
if(bin == null)
return null;
GATKBin gatkBin = new GATKBin(bin);
if(indexContent == null)
return null; return null;
final int binLevel = getLevelForBin(bin); final int binLevel = getLevelForBin(bin);
@ -312,16 +326,16 @@ public class GATKBAMIndex implements BAMIndex, BrowseableBAMIndex {
// Add the specified bin to the tree if it exists. // Add the specified bin to the tree if it exists.
List<GATKBin> binTree = new ArrayList<GATKBin>(); List<GATKBin> binTree = new ArrayList<GATKBin>();
if(indexQuery.containsBin(gatkBin)) if(indexContent.containsBin(gatkBin))
binTree.add(indexQuery.getBins().getBin(gatkBin.getBinNumber())); binTree.add(indexContent.getBins().getBin(gatkBin.getBinNumber()));
int currentBinLevel = binLevel; int currentBinLevel = binLevel;
while(--currentBinLevel >= 0) { while(--currentBinLevel >= 0) {
final int binStart = getFirstBinInLevel(currentBinLevel); final int binStart = getFirstBinInLevel(currentBinLevel);
final int binWidth = getMaxAddressibleGenomicLocation()/getLevelSize(currentBinLevel); final int binWidth = getMaxAddressibleGenomicLocation()/getLevelSize(currentBinLevel);
final int binNumber = firstLocusInBin/binWidth + binStart; final int binNumber = firstLocusInBin/binWidth + binStart;
GATKBin parentBin = indexQuery.getBins().getBin(binNumber); GATKBin parentBin = indexContent.getBins().getBin(binNumber);
if(parentBin != null && indexQuery.containsBin(parentBin)) if(parentBin != null && indexContent.containsBin(parentBin))
binTree.add(parentBin); binTree.add(parentBin);
} }
@ -332,7 +346,7 @@ public class GATKBAMIndex implements BAMIndex, BrowseableBAMIndex {
} }
final int start = getFirstLocusInBin(bin); final int start = getFirstLocusInBin(bin);
chunkList = optimizeChunkList(chunkList,indexQuery.getLinearIndex().getMinimumOffset(start)); chunkList = optimizeChunkList(chunkList,indexContent.getLinearIndex().getMinimumOffset(start));
return new GATKBAMFileSpan(chunkList.toArray(new GATKChunk[chunkList.size()])); return new GATKBAMFileSpan(chunkList.toArray(new GATKChunk[chunkList.size()]));
} }
@ -462,7 +476,7 @@ public class GATKBAMIndex implements BAMIndex, BrowseableBAMIndex {
* @param referenceIndex The reference to load. CachingBAMFileIndex only stores index data for entire references. * @param referenceIndex The reference to load. CachingBAMFileIndex only stores index data for entire references.
* @return The index information for this reference. * @return The index information for this reference.
*/ */
protected BAMIndexContent getQueryResults(final int referenceIndex) { public BAMIndexContent getQueryResults(final int referenceIndex) {
// If not in the cache, attempt to load it from disk. // If not in the cache, attempt to load it from disk.
return query(referenceIndex,1,-1); return query(referenceIndex,1,-1);
} }