GATK now uses a optimized indexing scheme in Tribble. 5x or more performance gain on files with many genotypes. Updated integrationtest that was failing and was clearly wrong. DB=; isn't a valid annotation.

git-svn-id: file:///humgen/gsa-scr1/gsa-engineering/svn_contents/trunk@3596 348d0f76-0448-11de-a6fe-93d51630548a

java/src/org/broadinstitute/sting/gatk/refdata/features/annotator/AnnotatorInputTableCodec.java

@@ -31,6 +31,7 @@ import java.io.IOException;
 import java.util.ArrayList;
 
 import org.apache.log4j.Logger;
+import org.broad.tribble.Feature;
 import org.broad.tribble.FeatureCodec;
 import org.broad.tribble.exception.CodecLineParsingException;
 import org.broad.tribble.util.AsciiLineReader;
@@ -74,7 +75,11 @@ public class AnnotatorInputTableCodec implements FeatureCodec<AnnotatorInputTabl
         return null;  // TODO: do we want the header to be a concrete type?
     }
 
-
+    // todo -- probably worth implementing for performance reasons
+    public Feature decodeLoc(String line) {
+        return decode(line);
+    }
+    
     /**
      * Parses the line into an AnnotatorInputTableFeature object.
      *

java/src/org/broadinstitute/sting/gatk/refdata/features/beagle/BeagleCodec.java

@@ -1,5 +1,6 @@
 package org.broadinstitute.sting.gatk.refdata.features.beagle;
 
+import org.broad.tribble.Feature;
 import org.broad.tribble.FeatureCodec;
 import org.broadinstitute.sting.gatk.refdata.features.annotator.AnnotatorInputTableFeature;
 /*
@@ -57,6 +58,10 @@ public class BeagleCodec implements FeatureCodec<BeagleFeature> {
 
     private static final String delimiterRegex = "\\s+";
 
+    public Feature decodeLoc(String line) {
+        return decode(line);
+    }
+    
     public static String[] readHeader(final File source) throws IOException {
         FileInputStream is = new FileInputStream(source);
         try {

java/src/org/broadinstitute/sting/gatk/refdata/features/sampileup/SAMPileupCodec.java

@@ -76,6 +76,10 @@ public class SAMPileupCodec implements FeatureCodec<SAMPileupFeature> {
         return null;  // we don't have a header
     }
 
+    public Feature decodeLoc(String line) {
+        return decode(line);
+    }
+
     public SAMPileupFeature decode(String line) {
 //       0          1             2         3                  4         5            6         7
 //*     chrX     466           T         Y                170      170       88      32 ... (piles of read bases  and quals follow)

java/src/org/broadinstitute/sting/gatk/refdata/features/samread/SAMReadCodec.java

@@ -24,6 +24,7 @@
 
 package org.broadinstitute.sting.gatk.refdata.features.samread;
 
+import org.broad.tribble.Feature;
 import org.broad.tribble.FeatureCodec;
 import org.broad.tribble.exception.CodecLineParsingException;
 import org.broad.tribble.util.ParsingUtils;
@@ -66,6 +67,10 @@ public class SAMReadCodec implements FeatureCodec<SAMReadFeature> {
         return null;  // we haven't stored the header
     }
 
+    public Feature decodeLoc(String line) {
+        return decode(line);
+    }
+
     /**
      * Decode a single line in a SAM text file.
      * @param line line to decode.

java/src/org/broadinstitute/sting/gatk/refdata/features/vcf4/VCF4Codec.java

@@ -135,6 +135,10 @@ public class VCF4Codec implements FeatureCodec {
     // a key optimization -- we need a per thread string parts array, so we don't allocate a big array over and over
     private String[] parts = null;
 
+    public Feature decodeLoc(String line) {
+        return decode(line);
+    }
+
     public Feature decode(String line) {
         if ( parts == null )
             parts = REQUIRE_HEADER ? new String[header.getColumnCount()] : new String[10000];    // todo -- remove require header

java/test/org/broadinstitute/sting/gatk/walkers/annotator/VariantAnnotatorIntegrationTest.java

@@ -122,7 +122,7 @@ public class VariantAnnotatorIntegrationTest extends WalkerTest {
     public void testDBTag() {
         WalkerTestSpec spec = new WalkerTestSpec(
                 baseTestString() + " -D " + GATKDataLocation + "dbsnp_129_b36.rod -G \"Standard\" -B variant,VCF," + validationDataLocation + "vcfexample3empty.vcf -BTI variant", 1,
-                Arrays.asList("05a9dcb87efc65e5c726c201460192d9"));
+                Arrays.asList("c3e0361af8e98bda1bc3a27260cb2c4a"));
         executeTest("getting DB tag", spec);
     }
-}
+}

settings/repository/org.broad/tribble-105.xml

@@ -1,3 +1,3 @@
 <ivy-module version="1.0">
-    <info organisation="org.broad" module="tribble" revision="104" status="integration" publication="201006171154200" />
+    <info organisation="org.broad" module="tribble" revision="105" status="integration" publication="201006181154200" />
 </ivy-module>