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

Bug fixes. Was ignoring the '$' character in a few places where I shouldn't have been. 

git-svn-id: file:///humgen/gsa-scr1/gsa-engineering/svn_contents/trunk@1504 348d0f76-0448-11de-a6fe-93d51630548a
This commit is contained in:
hanna 2009-09-02 16:27:31 +00:00
parent 55013eff78
commit d8aff9a925
1 changed files with 12 additions and 18 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

28
java/src/org/broadinstitute/sting/bwa/CreateBWTFromReference.java
View File

@ -30,13 +30,10 @@ import net.sf.picard.reference.ReferenceSequenceFileFactory;
import net.sf.picard.reference.ReferenceSequence; import net.sf.picard.reference.ReferenceSequence;
import net.sf.samtools.util.StringUtil; import net.sf.samtools.util.StringUtil;
import java.io.FileNotFoundException;
import java.io.IOException; import java.io.IOException;
import java.io.File; import java.io.File;
import java.util.TreeSet; import java.util.TreeSet;
import java.util.Comparator; import java.util.Comparator;
import java.util.Arrays;
import java.util.Collections;
/** /**
* Create a suffix array data structure. * Create a suffix array data structure.
@ -96,26 +93,23 @@ public class CreateBWTFromReference {
private int[] createCompressedSuffixArray( int[] suffixArray, int[] inverseSuffixArray ) { private int[] createCompressedSuffixArray( int[] suffixArray, int[] inverseSuffixArray ) {
int[] compressedSuffixArray = new int[suffixArray.length]; int[] compressedSuffixArray = new int[suffixArray.length];
compressedSuffixArray[0] = inverseSuffixArray[0]; compressedSuffixArray[0] = inverseSuffixArray[0];
for( int i = 1; i < suffixArray.length; i++ ) { for( int i = 1; i < suffixArray.length; i++ )
if( suffixArray[i]+1 < suffixArray.length ) compressedSuffixArray[i] = inverseSuffixArray[suffixArray[i]+1];
compressedSuffixArray[i] = inverseSuffixArray[suffixArray[i]+1];
}
return compressedSuffixArray; return compressedSuffixArray;
} }
private char[] createBWT( String sequence, int[] suffixArray ) { private char[] createBWT( String sequence, int[] suffixArray ) {
char[] bwt = new char[suffixArray.length]; char[] bwt = new char[suffixArray.length];
for( int i = 0; i < suffixArray.length; i++ ) { for( int i = 0; i < suffixArray.length; i++ ) {
int sequenceEnd = (suffixArray[i] + suffixArray.length - 1) % suffixArray.length; // Find the first character after the current character in the rotation. If the character is past the end
if( sequenceEnd < sequence.length() ) // (in other words, '$'), back up to the previous character.
bwt[i] = sequence.charAt(sequenceEnd); int sequenceEnd = Math.min((suffixArray[i]+suffixArray.length-1)%suffixArray.length, sequence.length()-1 );
else bwt[i] = sequence.charAt(sequenceEnd);
bwt[i] = '$';
} }
return bwt; return bwt;
} }
public static void main( String argv[] ) throws FileNotFoundException, IOException { public static void main( String argv[] ) throws IOException {
if( argv.length != 1 ) { if( argv.length != 1 ) {
System.out.println("No reference"); System.out.println("No reference");
return; return;
@ -134,18 +128,18 @@ public class CreateBWTFromReference {
// Generate the suffix array and print diagnostics. // Generate the suffix array and print diagnostics.
int[] suffixArray = creator.createSuffixArray(sequence); int[] suffixArray = creator.createSuffixArray(sequence);
for( int i = 0; i < 10; i++ ) for( int i = 0; i < 8; i++ )
System.out.printf("suffixArray[%d] = %d (%s...)%n", i, suffixArray[i], sequence.substring(suffixArray[i],Math.min(suffixArray[i]+100,sequence.length()))); System.out.printf("suffixArray[%d] = %d (%s...)%n", i, suffixArray[i], sequence.substring(suffixArray[i],Math.min(suffixArray[i]+100,sequence.length())));
// Invert the suffix array and print diagnostics. // Invert the suffix array and print diagnostics.
int[] inverseSuffixArray = creator.invertSuffixArray(suffixArray); int[] inverseSuffixArray = creator.invertSuffixArray(suffixArray);
for( int i = 0; i < 10; i++ ) for( int i = 0; i < 8; i++ )
System.out.printf("inverseSuffixArray[%d] = %d (%s...)%n", i, inverseSuffixArray[i], sequence.substring(i,Math.min(i+100,sequence.length()))); System.out.printf("inverseSuffixArray[%d] = %d (%s...)%n", i, inverseSuffixArray[i], sequence.substring(i,Math.min(i+100,sequence.length())));
// Create the data structure for the compressed suffix array and print diagnostics. // Create the data structure for the compressed suffix array and print diagnostics.
int[] compressedSuffixArray = creator.createCompressedSuffixArray(suffixArray,inverseSuffixArray); int[] compressedSuffixArray = creator.createCompressedSuffixArray(suffixArray,inverseSuffixArray);
int reconstructedInverseSA = compressedSuffixArray[0]; int reconstructedInverseSA = compressedSuffixArray[0];
for( int i = 0; i < 10; i++ ) { for( int i = 0; i < 8; i++ ) {
System.out.printf("compressedSuffixArray[%d] = %d (SA-1[%d] = %d)%n", i, compressedSuffixArray[i], i, reconstructedInverseSA); System.out.printf("compressedSuffixArray[%d] = %d (SA-1[%d] = %d)%n", i, compressedSuffixArray[i], i, reconstructedInverseSA);
reconstructedInverseSA = compressedSuffixArray[reconstructedInverseSA]; reconstructedInverseSA = compressedSuffixArray[reconstructedInverseSA];
} }
@ -167,7 +161,7 @@ public class CreateBWTFromReference {
/** /**
* Create a new comparator. * Create a new comparator.
* @param sequence * @param sequence Reference sequence to use as basis for comparison.
*/ */
public SuffixArrayComparator( String sequence ) { public SuffixArrayComparator( String sequence ) {
this.sequence = sequence; this.sequence = sequence;