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

Merge pull request #554 from broadinstitute/bh_SOR_new_annotation 

Bh sor new annotation
This commit is contained in:
Eric Banks 2014-03-17 10:58:13 -04:00
parent bb4a7bf87e 82108d110f
commit 34c697bf12
4 changed files with 386 additions and 79 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

87
protected/gatk-protected/src/main/java/org/broadinstitute/sting/gatk/walkers/annotator/FisherStrand.java
View File

@ -53,13 +53,10 @@ import org.broadinstitute.sting.gatk.contexts.ReferenceContext;
import org.broadinstitute.sting.gatk.refdata.RefMetaDataTracker;
import org.broadinstitute.sting.gatk.walkers.annotator.interfaces.ActiveRegionBasedAnnotation;
import org.broadinstitute.sting.gatk.walkers.annotator.interfaces.AnnotatorCompatible;
import org.broadinstitute.sting.gatk.walkers.annotator.interfaces.InfoFieldAnnotation;
import org.broadinstitute.sting.gatk.walkers.annotator.interfaces.StandardAnnotation;
import org.broadinstitute.sting.utils.genotyper.MostLikelyAllele;
import org.broadinstitute.sting.utils.genotyper.PerReadAlleleLikelihoodMap;
import org.broadinstitute.sting.utils.QualityUtils;
import org.broadinstitute.variant.variantcontext.Genotype;
import org.broadinstitute.variant.variantcontext.GenotypesContext;
import org.broadinstitute.variant.vcf.VCFHeaderLineType;
import org.broadinstitute.variant.vcf.VCFInfoHeaderLine;
import org.broadinstitute.sting.utils.pileup.PileupElement;
@ -81,7 +78,7 @@ import java.util.*;
* <h3>Caveat</h3>
* <p>The Fisher Strand test may not be calculated for certain complex indel cases or for multi-allelic sites.</p>
*/
public class FisherStrand extends InfoFieldAnnotation implements StandardAnnotation, ActiveRegionBasedAnnotation {
public class FisherStrand extends StrandBiasTest implements StandardAnnotation, ActiveRegionBasedAnnotation {
private final static boolean ENABLE_DEBUGGING = false;
private final static Logger logger = Logger.getLogger(FisherStrand.class);
@ -100,7 +97,7 @@ public class FisherStrand extends InfoFieldAnnotation implements StandardAnnotat
return null;
if ( vc.hasGenotypes() ) {
final int[][] tableFromPerSampleAnnotations = getTableFromSamples( vc.getGenotypes() );
final int[][] tableFromPerSampleAnnotations = getTableFromSamples( vc.getGenotypes(), MIN_COUNT );
if ( tableFromPerSampleAnnotations != null ) {
return pValueForBestTable(tableFromPerSampleAnnotations, null);
}
@ -116,8 +113,8 @@ public class FisherStrand extends InfoFieldAnnotation implements StandardAnnotat
else if (stratifiedPerReadAlleleLikelihoodMap != null) {
// either SNP with no alignment context, or indels: per-read likelihood map needed
final int[][] table = getContingencyTable(stratifiedPerReadAlleleLikelihoodMap, vc);
// logger.info("VC " + vc);
// printTable(table, 0.0);
//logger.info("VC " + vc);
//printTable(table, 0.0);
return pValueForBestTable(table, null);
}
else
@ -126,45 +123,6 @@ public class FisherStrand extends InfoFieldAnnotation implements StandardAnnotat
return null;
}
/**
* Create the FisherStrand table by retrieving the per-sample strand bias annotation and adding them together
* @param genotypes the genotypes from which to pull out the per-sample strand bias annotation
* @return the table used for the FisherStrand p-value calculation, will be null if none of the genotypes contain the per-sample SB annotation
*/
private int[][] getTableFromSamples( final GenotypesContext genotypes ) {
if( genotypes == null ) { throw new IllegalArgumentException("Genotypes cannot be null."); }
final int[] sbArray = {0,0,0,0}; // reference-forward-reverse -by- alternate-forward-reverse
boolean foundData = false;
for( final Genotype g : genotypes ) {
if( g.isNoCall() || ! g.hasAnyAttribute(StrandBiasBySample.STRAND_BIAS_BY_SAMPLE_KEY_NAME) )
continue;
foundData = true;
final String sbbsString = (String) g.getAnyAttribute(StrandBiasBySample.STRAND_BIAS_BY_SAMPLE_KEY_NAME);
final int[] data = encodeSBBS(sbbsString);
if ( passesMinimumThreshold(data) ) {
for( int index = 0; index < sbArray.length; index++ ) {
sbArray[index] += data[index];
}
}
}
return ( foundData ? decodeSBBS(sbArray) : null );
}
/**
* Does this strand data array pass the minimum threshold for inclusion?
*
* @param data the array
* @return true if it passes the minimum threshold, false otherwise
*/
private static boolean passesMinimumThreshold(final int[] data) {
// the ref and alt totals must each be greater than MIN_COUNT
return data[0] + data[1] > MIN_COUNT && data[2] + data[3] > MIN_COUNT;
}
/**
* Create an annotation for the highest (i.e., least significant) p-value of table1 and table2
*
@ -190,7 +148,7 @@ public class FisherStrand extends InfoFieldAnnotation implements StandardAnnotat
* @param pValue
* @return a hash map from FS -> phred-scaled pValue
*/
private Map<String, Object> annotationForOneTable(final double pValue) {
protected Map<String, Object> annotationForOneTable(final double pValue) {
final Object value = String.format("%.3f", QualityUtils.phredScaleErrorRate(Math.max(pValue, MIN_PVALUE))); // prevent INFINITYs
return Collections.singletonMap(FS, value);
}
@ -218,36 +176,6 @@ public class FisherStrand extends InfoFieldAnnotation implements StandardAnnotat
list.add(table[1][1]);
return list;
}
/**
* Helper function to parse the genotype annotation into the SB annotation array
* @param string the string that is returned by genotype.getAnnotation("SB")
* @return the array used by the per-sample Strand Bias annotation
*/
private static int[] encodeSBBS( final String string ) {
final int[] array = new int[4];
final StringTokenizer tokenizer = new StringTokenizer(string, ",", false);
for( int index = 0; index < 4; index++ ) {
array[index] = Integer.parseInt(tokenizer.nextToken());
}
return array;
}
/**
* Helper function to turn the SB annotation array into the FisherStrand table
* @param array the array used by the per-sample Strand Bias annotation
* @return the table used by the FisherStrand annotation
*/
private static int[][] decodeSBBS( final int[] array ) {
if(array.length != 4) { throw new IllegalArgumentException("Expecting a length = 4 strand bias array."); }
final int[][] table = new int[2][2];
table[0][0] = array[0];
table[0][1] = array[1];
table[1][0] = array[2];
table[1][1] = array[3];
return table;
}
private Double pValueForContingencyTable(int[][] originalTable) {
final int[][] normalizedTable = normalizeContingencyTable(originalTable);
@ -419,7 +347,7 @@ public class FisherStrand extends InfoFieldAnnotation implements StandardAnnotat
final GATKSAMRecord read = el.getKey();
updateTable(myTable, mostLikelyAllele.getAlleleIfInformative(), read, ref, alt);
}
if ( passesMinimumThreshold(myTable) )
if ( passesMinimumThreshold(myTable, MIN_COUNT) )
copyToMainTable(myTable, table);
}
@ -464,7 +392,8 @@ public class FisherStrand extends InfoFieldAnnotation implements StandardAnnotat
updateTable(myTable, Allele.create(p.getBase(), false), p.getRead(), ref, alt);
}
if ( passesMinimumThreshold(myTable) )
if ( passesMinimumThreshold( myTable, MIN_COUNT ) )
copyToMainTable(myTable, table);
}

128
protected/gatk-protected/src/main/java/org/broadinstitute/sting/gatk/walkers/annotator/StrandBiasTest.java 100644
View File

@ -0,0 +1,128 @@
/*
* By downloading the PROGRAM you agree to the following terms of use:
*
* BROAD INSTITUTE - SOFTWARE LICENSE AGREEMENT - FOR ACADEMIC NON-COMMERCIAL RESEARCH PURPOSES ONLY
*
* This Agreement is made between the Broad Institute, Inc. with a principal address at 7 Cambridge Center, Cambridge, MA 02142 (BROAD) and the LICENSEE and is effective at the date the downloading is completed (EFFECTIVE DATE).
*
* WHEREAS, LICENSEE desires to license the PROGRAM, as defined hereinafter, and BROAD wishes to have this PROGRAM utilized in the public interest, subject only to the royalty-free, nonexclusive, nontransferable license rights of the United States Government pursuant to 48 CFR 52.227-14; and
* WHEREAS, LICENSEE desires to license the PROGRAM and BROAD desires to grant a license on the following terms and conditions.
* NOW, THEREFORE, in consideration of the promises and covenants made herein, the parties hereto agree as follows:
*
* 1. DEFINITIONS
* 1.1 PROGRAM shall mean copyright in the object code and source code known as GATK2 and related documentation, if any, as they exist on the EFFECTIVE DATE and can be downloaded from http://www.broadinstitute/GATK on the EFFECTIVE DATE.
*
* 2. LICENSE
* 2.1 Grant. Subject to the terms of this Agreement, BROAD hereby grants to LICENSEE, solely for academic non-commercial research purposes, a non-exclusive, non-transferable license to: (a) download, execute and display the PROGRAM and (b) create bug fixes and modify the PROGRAM.
* The LICENSEE may apply the PROGRAM in a pipeline to data owned by users other than the LICENSEE and provide these users the results of the PROGRAM provided LICENSEE does so for academic non-commercial purposes only. For clarification purposes, academic sponsored research is not a commercial use under the terms of this Agreement.
* 2.2 No Sublicensing or Additional Rights. LICENSEE shall not sublicense or distribute the PROGRAM, in whole or in part, without prior written permission from BROAD. LICENSEE shall ensure that all of its users agree to the terms of this Agreement. LICENSEE further agrees that it shall not put the PROGRAM on a network, server, or other similar technology that may be accessed by anyone other than the LICENSEE and its employees and users who have agreed to the terms of this agreement.
* 2.3 License Limitations. Nothing in this Agreement shall be construed to confer any rights upon LICENSEE by implication, estoppel, or otherwise to any computer software, trademark, intellectual property, or patent rights of BROAD, or of any other entity, except as expressly granted herein. LICENSEE agrees that the PROGRAM, in whole or part, shall not be used for any commercial purpose, including without limitation, as the basis of a commercial software or hardware product or to provide services. LICENSEE further agrees that the PROGRAM shall not be copied or otherwise adapted in order to circumvent the need for obtaining a license for use of the PROGRAM.
*
* 3. OWNERSHIP OF INTELLECTUAL PROPERTY
* LICENSEE acknowledges that title to the PROGRAM shall remain with BROAD. The PROGRAM is marked with the following BROAD copyright notice and notice of attribution to contributors. LICENSEE shall retain such notice on all copies. LICENSEE agrees to include appropriate attribution if any results obtained from use of the PROGRAM are included in any publication.
* Copyright 2012 Broad Institute, Inc.
* Notice of attribution: The GATK2 program was made available through the generosity of Medical and Population Genetics program at the Broad Institute, Inc.
* LICENSEE shall not use any trademark or trade name of BROAD, or any variation, adaptation, or abbreviation, of such marks or trade names, or any names of officers, faculty, students, employees, or agents of BROAD except as states above for attribution purposes.
*
* 4. INDEMNIFICATION
* LICENSEE shall indemnify, defend, and hold harmless BROAD, and their respective officers, faculty, students, employees, associated investigators and agents, and their respective successors, heirs and assigns, (Indemnitees), against any liability, damage, loss, or expense (including reasonable attorneys fees and expenses) incurred by or imposed upon any of the Indemnitees in connection with any claims, suits, actions, demands or judgments arising out of any theory of liability (including, without limitation, actions in the form of tort, warranty, or strict liability and regardless of whether such action has any factual basis) pursuant to any right or license granted under this Agreement.
*
* 5. NO REPRESENTATIONS OR WARRANTIES
* THE PROGRAM IS DELIVERED AS IS. BROAD MAKES NO REPRESENTATIONS OR WARRANTIES OF ANY KIND CONCERNING THE PROGRAM OR THE COPYRIGHT, EXPRESS OR IMPLIED, INCLUDING, WITHOUT LIMITATION, WARRANTIES OF MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE, NONINFRINGEMENT, OR THE ABSENCE OF LATENT OR OTHER DEFECTS, WHETHER OR NOT DISCOVERABLE. BROAD EXTENDS NO WARRANTIES OF ANY KIND AS TO PROGRAM CONFORMITY WITH WHATEVER USER MANUALS OR OTHER LITERATURE MAY BE ISSUED FROM TIME TO TIME.
* IN NO EVENT SHALL BROAD OR ITS RESPECTIVE DIRECTORS, OFFICERS, EMPLOYEES, AFFILIATED INVESTIGATORS AND AFFILIATES BE LIABLE FOR INCIDENTAL OR CONSEQUENTIAL DAMAGES OF ANY KIND, INCLUDING, WITHOUT LIMITATION, ECONOMIC DAMAGES OR INJURY TO PROPERTY AND LOST PROFITS, REGARDLESS OF WHETHER BROAD SHALL BE ADVISED, SHALL HAVE OTHER REASON TO KNOW, OR IN FACT SHALL KNOW OF THE POSSIBILITY OF THE FOREGOING.
*
* 6. ASSIGNMENT
* This Agreement is personal to LICENSEE and any rights or obligations assigned by LICENSEE without the prior written consent of BROAD shall be null and void.
*
* 7. MISCELLANEOUS
* 7.1 Export Control. LICENSEE gives assurance that it will comply with all United States export control laws and regulations controlling the export of the PROGRAM, including, without limitation, all Export Administration Regulations of the United States Department of Commerce. Among other things, these laws and regulations prohibit, or require a license for, the export of certain types of software to specified countries.
* 7.2 Termination. LICENSEE shall have the right to terminate this Agreement for any reason upon prior written notice to BROAD. If LICENSEE breaches any provision hereunder, and fails to cure such breach within thirty (30) days, BROAD may terminate this Agreement immediately. Upon termination, LICENSEE shall provide BROAD with written assurance that the original and all copies of the PROGRAM have been destroyed, except that, upon prior written authorization from BROAD, LICENSEE may retain a copy for archive purposes.
* 7.3 Survival. The following provisions shall survive the expiration or termination of this Agreement: Articles 1, 3, 4, 5 and Sections 2.2, 2.3, 7.3, and 7.4.
* 7.4 Notice. Any notices under this Agreement shall be in writing, shall specifically refer to this Agreement, and shall be sent by hand, recognized national overnight courier, confirmed facsimile transmission, confirmed electronic mail, or registered or certified mail, postage prepaid, return receipt requested. All notices under this Agreement shall be deemed effective upon receipt.
* 7.5 Amendment and Waiver; Entire Agreement. This Agreement may be amended, supplemented, or otherwise modified only by means of a written instrument signed by all parties. Any waiver of any rights or failure to act in a specific instance shall relate only to such instance and shall not be construed as an agreement to waive any rights or fail to act in any other instance, whether or not similar. This Agreement constitutes the entire agreement among the parties with respect to its subject matter and supersedes prior agreements or understandings between the parties relating to its subject matter.
* 7.6 Binding Effect; Headings. This Agreement shall be binding upon and inure to the benefit of the parties and their respective permitted successors and assigns. All headings are for convenience only and shall not affect the meaning of any provision of this Agreement.
* 7.7 Governing Law. This Agreement shall be construed, governed, interpreted and applied in accordance with the internal laws of the Commonwealth of Massachusetts, U.S.A., without regard to conflict of laws principles.
*/
package org.broadinstitute.sting.gatk.walkers.annotator;
import org.broadinstitute.sting.gatk.walkers.annotator.interfaces.InfoFieldAnnotation;
import org.broadinstitute.variant.variantcontext.Genotype;
import org.broadinstitute.variant.variantcontext.GenotypesContext;
import java.util.*;
/**
* Class of tests to detect strand bias.
*/
public abstract class StrandBiasTest extends InfoFieldAnnotation {
/**
* Create the contingency table by retrieving the per-sample strand bias annotation and adding them together
* @param genotypes the genotypes from which to pull out the per-sample strand bias annotation
* @param minCount minimum threshold for the sample strand bias counts for each ref and alt.
* If both ref and alt counts are above minCount the whole sample strand bias is added to the resulting table
* @return the table used for several strand bias tests, will be null if none of the genotypes contain the per-sample SB annotation
*/
protected int[][] getTableFromSamples( final GenotypesContext genotypes, final int minCount ) {
if( genotypes == null ) { throw new IllegalArgumentException("Genotypes cannot be null."); }
final int[] sbArray = {0,0,0,0}; // reference-forward-reverse -by- alternate-forward-reverse
boolean foundData = false;
for( final Genotype g : genotypes ) {
if( g.isNoCall() || ! g.hasAnyAttribute(StrandBiasBySample.STRAND_BIAS_BY_SAMPLE_KEY_NAME) )
continue;
foundData = true;
final String sbbsString = (String) g.getAnyAttribute(StrandBiasBySample.STRAND_BIAS_BY_SAMPLE_KEY_NAME);
final int[] data = encodeSBBS(sbbsString);
if ( passesMinimumThreshold(data, minCount) ) {
for( int index = 0; index < sbArray.length; index++ ) {
sbArray[index] += data[index];
}
}
}
return ( foundData ? decodeSBBS(sbArray) : null );
}
/**
* Does this strand data array pass the minimum threshold for inclusion?
*
* @param data the array
* @minCount The minimum threshold of counts in the array
* @return true if it passes the minimum threshold, false otherwise
*/
protected static boolean passesMinimumThreshold(final int[] data, final int minCount) {
// the ref and alt totals must each be greater than MIN_COUNT
return data[0] + data[1] > minCount && data[2] + data[3] > minCount;
}
/**
* Helper function to parse the genotype annotation into the SB annotation array
* @param string the string that is returned by genotype.getAnnotation("SB")
* @return the array used by the per-sample Strand Bias annotation
*/
private static int[] encodeSBBS( final String string ) {
final int[] array = new int[4];
final StringTokenizer tokenizer = new StringTokenizer(string, ",", false);
for( int index = 0; index < 4; index++ ) {
array[index] = Integer.parseInt(tokenizer.nextToken());
}
return array;
}
/**
* Helper function to turn the SB annotation array into a contingency table
* @param array the array used by the per-sample Strand Bias annotation
* @return the table used by the StrandOddsRatio annotation
*/
private static int[][] decodeSBBS( final int[] array ) {
if(array.length != 4) { throw new IllegalArgumentException("Expecting a length = 4 strand bias array."); }
final int[][] table = new int[2][2];
table[0][0] = array[0];
table[0][1] = array[1];
table[1][0] = array[2];
table[1][1] = array[3];
return table;
}
}

150
protected/gatk-protected/src/main/java/org/broadinstitute/sting/gatk/walkers/annotator/StrandOddsRatio.java 100644
View File

@ -0,0 +1,150 @@
/*
* By downloading the PROGRAM you agree to the following terms of use:
*
* BROAD INSTITUTE - SOFTWARE LICENSE AGREEMENT - FOR ACADEMIC NON-COMMERCIAL RESEARCH PURPOSES ONLY
*
* This Agreement is made between the Broad Institute, Inc. with a principal address at 7 Cambridge Center, Cambridge, MA 02142 (BROAD) and the LICENSEE and is effective at the date the downloading is completed (EFFECTIVE DATE).
*
* WHEREAS, LICENSEE desires to license the PROGRAM, as defined hereinafter, and BROAD wishes to have this PROGRAM utilized in the public interest, subject only to the royalty-free, nonexclusive, nontransferable license rights of the United States Government pursuant to 48 CFR 52.227-14; and
* WHEREAS, LICENSEE desires to license the PROGRAM and BROAD desires to grant a license on the following terms and conditions.
* NOW, THEREFORE, in consideration of the promises and covenants made herein, the parties hereto agree as follows:
*
* 1. DEFINITIONS
* 1.1 PROGRAM shall mean copyright in the object code and source code known as GATK2 and related documentation, if any, as they exist on the EFFECTIVE DATE and can be downloaded from http://www.broadinstitute/GATK on the EFFECTIVE DATE.
*
* 2. LICENSE
* 2.1 Grant. Subject to the terms of this Agreement, BROAD hereby grants to LICENSEE, solely for academic non-commercial research purposes, a non-exclusive, non-transferable license to: (a) download, execute and display the PROGRAM and (b) create bug fixes and modify the PROGRAM.
* The LICENSEE may apply the PROGRAM in a pipeline to data owned by users other than the LICENSEE and provide these users the results of the PROGRAM provided LICENSEE does so for academic non-commercial purposes only. For clarification purposes, academic sponsored research is not a commercial use under the terms of this Agreement.
* 2.2 No Sublicensing or Additional Rights. LICENSEE shall not sublicense or distribute the PROGRAM, in whole or in part, without prior written permission from BROAD. LICENSEE shall ensure that all of its users agree to the terms of this Agreement. LICENSEE further agrees that it shall not put the PROGRAM on a network, server, or other similar technology that may be accessed by anyone other than the LICENSEE and its employees and users who have agreed to the terms of this agreement.
* 2.3 License Limitations. Nothing in this Agreement shall be construed to confer any rights upon LICENSEE by implication, estoppel, or otherwise to any computer software, trademark, intellectual property, or patent rights of BROAD, or of any other entity, except as expressly granted herein. LICENSEE agrees that the PROGRAM, in whole or part, shall not be used for any commercial purpose, including without limitation, as the basis of a commercial software or hardware product or to provide services. LICENSEE further agrees that the PROGRAM shall not be copied or otherwise adapted in order to circumvent the need for obtaining a license for use of the PROGRAM.
*
* 3. OWNERSHIP OF INTELLECTUAL PROPERTY
* LICENSEE acknowledges that title to the PROGRAM shall remain with BROAD. The PROGRAM is marked with the following BROAD copyright notice and notice of attribution to contributors. LICENSEE shall retain such notice on all copies. LICENSEE agrees to include appropriate attribution if any results obtained from use of the PROGRAM are included in any publication.
* Copyright 2012 Broad Institute, Inc.
* Notice of attribution: The GATK2 program was made available through the generosity of Medical and Population Genetics program at the Broad Institute, Inc.
* LICENSEE shall not use any trademark or trade name of BROAD, or any variation, adaptation, or abbreviation, of such marks or trade names, or any names of officers, faculty, students, employees, or agents of BROAD except as states above for attribution purposes.
*
* 4. INDEMNIFICATION
* LICENSEE shall indemnify, defend, and hold harmless BROAD, and their respective officers, faculty, students, employees, associated investigators and agents, and their respective successors, heirs and assigns, (Indemnitees), against any liability, damage, loss, or expense (including reasonable attorneys fees and expenses) incurred by or imposed upon any of the Indemnitees in connection with any claims, suits, actions, demands or judgments arising out of any theory of liability (including, without limitation, actions in the form of tort, warranty, or strict liability and regardless of whether such action has any factual basis) pursuant to any right or license granted under this Agreement.
*
* 5. NO REPRESENTATIONS OR WARRANTIES
* THE PROGRAM IS DELIVERED AS IS. BROAD MAKES NO REPRESENTATIONS OR WARRANTIES OF ANY KIND CONCERNING THE PROGRAM OR THE COPYRIGHT, EXPRESS OR IMPLIED, INCLUDING, WITHOUT LIMITATION, WARRANTIES OF MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE, NONINFRINGEMENT, OR THE ABSENCE OF LATENT OR OTHER DEFECTS, WHETHER OR NOT DISCOVERABLE. BROAD EXTENDS NO WARRANTIES OF ANY KIND AS TO PROGRAM CONFORMITY WITH WHATEVER USER MANUALS OR OTHER LITERATURE MAY BE ISSUED FROM TIME TO TIME.
* IN NO EVENT SHALL BROAD OR ITS RESPECTIVE DIRECTORS, OFFICERS, EMPLOYEES, AFFILIATED INVESTIGATORS AND AFFILIATES BE LIABLE FOR INCIDENTAL OR CONSEQUENTIAL DAMAGES OF ANY KIND, INCLUDING, WITHOUT LIMITATION, ECONOMIC DAMAGES OR INJURY TO PROPERTY AND LOST PROFITS, REGARDLESS OF WHETHER BROAD SHALL BE ADVISED, SHALL HAVE OTHER REASON TO KNOW, OR IN FACT SHALL KNOW OF THE POSSIBILITY OF THE FOREGOING.
*
* 6. ASSIGNMENT
* This Agreement is personal to LICENSEE and any rights or obligations assigned by LICENSEE without the prior written consent of BROAD shall be null and void.
*
* 7. MISCELLANEOUS
* 7.1 Export Control. LICENSEE gives assurance that it will comply with all United States export control laws and regulations controlling the export of the PROGRAM, including, without limitation, all Export Administration Regulations of the United States Department of Commerce. Among other things, these laws and regulations prohibit, or require a license for, the export of certain types of software to specified countries.
* 7.2 Termination. LICENSEE shall have the right to terminate this Agreement for any reason upon prior written notice to BROAD. If LICENSEE breaches any provision hereunder, and fails to cure such breach within thirty (30) days, BROAD may terminate this Agreement immediately. Upon termination, LICENSEE shall provide BROAD with written assurance that the original and all copies of the PROGRAM have been destroyed, except that, upon prior written authorization from BROAD, LICENSEE may retain a copy for archive purposes.
* 7.3 Survival. The following provisions shall survive the expiration or termination of this Agreement: Articles 1, 3, 4, 5 and Sections 2.2, 2.3, 7.3, and 7.4.
* 7.4 Notice. Any notices under this Agreement shall be in writing, shall specifically refer to this Agreement, and shall be sent by hand, recognized national overnight courier, confirmed facsimile transmission, confirmed electronic mail, or registered or certified mail, postage prepaid, return receipt requested. All notices under this Agreement shall be deemed effective upon receipt.
* 7.5 Amendment and Waiver; Entire Agreement. This Agreement may be amended, supplemented, or otherwise modified only by means of a written instrument signed by all parties. Any waiver of any rights or failure to act in a specific instance shall relate only to such instance and shall not be construed as an agreement to waive any rights or fail to act in any other instance, whether or not similar. This Agreement constitutes the entire agreement among the parties with respect to its subject matter and supersedes prior agreements or understandings between the parties relating to its subject matter.
* 7.6 Binding Effect; Headings. This Agreement shall be binding upon and inure to the benefit of the parties and their respective permitted successors and assigns. All headings are for convenience only and shall not affect the meaning of any provision of this Agreement.
* 7.7 Governing Law. This Agreement shall be construed, governed, interpreted and applied in accordance with the internal laws of the Commonwealth of Massachusetts, U.S.A., without regard to conflict of laws principles.
*/
package org.broadinstitute.sting.gatk.walkers.annotator;
import org.broadinstitute.sting.gatk.contexts.AlignmentContext;
import org.broadinstitute.sting.gatk.contexts.ReferenceContext;
import org.broadinstitute.sting.gatk.refdata.RefMetaDataTracker;
import org.broadinstitute.sting.gatk.walkers.annotator.interfaces.AnnotatorCompatible;
import org.broadinstitute.sting.gatk.walkers.annotator.interfaces.ActiveRegionBasedAnnotation;
import org.broadinstitute.sting.utils.genotyper.PerReadAlleleLikelihoodMap;
import org.broadinstitute.variant.variantcontext.VariantContext;
import org.broadinstitute.variant.vcf.VCFHeaderLineType;
import org.broadinstitute.variant.vcf.VCFInfoHeaderLine;
import java.util.*;
/**
* Symmetric Odds Ratio to detect strand bias
*
* <p> Odds Ratios in the 2x2 contingency table below are R = (X[0][0] * X[1][1]) / (X[0][1] * X[1][0]) and its inverse
* + strand - strand
* Ref X[0][0] X[0][1]
* Alt X[1][0] X[1][0]
* The sum R + 1/R is used to detect a difference in strand bias for ref and for alt (the sum makes it symmetric):
* A high value is indicative of large difference where one entry is very small compared to the others.
* </p>
*/
public class StrandOddsRatio extends StrandBiasTest implements ActiveRegionBasedAnnotation {
private final static double AUGMENTATION_CONSTANT = 0.1;
private static final int MIN_COUNT = 0;
private static final String SOR = "SOR";
public Map<String, Object> annotate(final RefMetaDataTracker tracker,
final AnnotatorCompatible walker,
final ReferenceContext ref,
final Map<String,AlignmentContext> stratifiedContexts,
final VariantContext vc,
final Map<String, PerReadAlleleLikelihoodMap> stratifiedPerReadAlleleLikelihoodMap) {
if ( !vc.isVariant() )
return null;
if ( vc.hasGenotypes() ) {
final int[][] tableFromPerSampleAnnotations = getTableFromSamples( vc.getGenotypes(), MIN_COUNT );
if ( tableFromPerSampleAnnotations != null ) {
final double ratio = symmetricOddsRatio(tableFromPerSampleAnnotations);
return annotationForOneTable(ratio);
}
}
return null;
}
/**
* Computes the symmetric odds ratio of a table after augmentation.
* Augmentation avoids quotient by zero.
*
* @param originalTable The table before augmentation
* @return the symmetric odds ratio
*/
final protected double symmetricOddsRatio(final int[][] originalTable) {
final double[][] augmentedTable = augmentContingencyTable(originalTable);
double ratio = 0;
ratio += (augmentedTable[0][0] / augmentedTable[0][1]) * (augmentedTable[1][1] / augmentedTable[1][0]);
ratio += (augmentedTable[0][1] / augmentedTable[0][0]) * (augmentedTable[1][0] / augmentedTable[1][1]);
return ratio;
}
/**
* Adds the small value AUGMENTATION_CONSTANT to all the entries of the table.
*
* @param table the table to augment
* @return the augmented table
*/
private static double[][] augmentContingencyTable(final int[][] table) {
double[][] augmentedTable = new double[2][2];
for ( int i = 0; i < 2; i++ ) {
for ( int j = 0; j < 2; j++ )
augmentedTable[i][j] = table[i][j] + AUGMENTATION_CONSTANT;
}
return augmentedTable;
}
/**
* Returns an annotation result given a ratio
*
* @param ratio the symmetric odds ratio of the contingency table
* @return a hash map from SOR
*/
protected Map<String, Object> annotationForOneTable(final double ratio) {
final Object value = String.format("%.3f", ratio);
return Collections.singletonMap(SOR, value);
}
public List<VCFInfoHeaderLine> getDescriptions() {
return Collections.singletonList(new VCFInfoHeaderLine(SOR, 1, VCFHeaderLineType.Float, "Symmetric Odds Ratio of 2x2 contingency table to detect strand bias"));
}
public List<String> getKeyNames() {
return Collections.singletonList(SOR);
}
}

100
protected/gatk-protected/src/test/java/org/broadinstitute/sting/gatk/walkers/annotator/StrandOddsRatioUnitTest.java 100644
View File

@ -0,0 +1,100 @@
/*
* By downloading the PROGRAM you agree to the following terms of use:
*
* BROAD INSTITUTE - SOFTWARE LICENSE AGREEMENT - FOR ACADEMIC NON-COMMERCIAL RESEARCH PURPOSES ONLY
*
* This Agreement is made between the Broad Institute, Inc. with a principal address at 7 Cambridge Center, Cambridge, MA 02142 (BROAD) and the LICENSEE and is effective at the date the downloading is completed (EFFECTIVE DATE).
*
* WHEREAS, LICENSEE desires to license the PROGRAM, as defined hereinafter, and BROAD wishes to have this PROGRAM utilized in the public interest, subject only to the royalty-free, nonexclusive, nontransferable license rights of the United States Government pursuant to 48 CFR 52.227-14; and
* WHEREAS, LICENSEE desires to license the PROGRAM and BROAD desires to grant a license on the following terms and conditions.
* NOW, THEREFORE, in consideration of the promises and covenants made herein, the parties hereto agree as follows:
*
* 1. DEFINITIONS
* 1.1 PROGRAM shall mean copyright in the object code and source code known as GATK2 and related documentation, if any, as they exist on the EFFECTIVE DATE and can be downloaded from http://www.broadinstitute/GATK on the EFFECTIVE DATE.
*
* 2. LICENSE
* 2.1 Grant. Subject to the terms of this Agreement, BROAD hereby grants to LICENSEE, solely for academic non-commercial research purposes, a non-exclusive, non-transferable license to: (a) download, execute and display the PROGRAM and (b) create bug fixes and modify the PROGRAM.
* The LICENSEE may apply the PROGRAM in a pipeline to data owned by users other than the LICENSEE and provide these users the results of the PROGRAM provided LICENSEE does so for academic non-commercial purposes only. For clarification purposes, academic sponsored research is not a commercial use under the terms of this Agreement.
* 2.2 No Sublicensing or Additional Rights. LICENSEE shall not sublicense or distribute the PROGRAM, in whole or in part, without prior written permission from BROAD. LICENSEE shall ensure that all of its users agree to the terms of this Agreement. LICENSEE further agrees that it shall not put the PROGRAM on a network, server, or other similar technology that may be accessed by anyone other than the LICENSEE and its employees and users who have agreed to the terms of this agreement.
* 2.3 License Limitations. Nothing in this Agreement shall be construed to confer any rights upon LICENSEE by implication, estoppel, or otherwise to any computer software, trademark, intellectual property, or patent rights of BROAD, or of any other entity, except as expressly granted herein. LICENSEE agrees that the PROGRAM, in whole or part, shall not be used for any commercial purpose, including without limitation, as the basis of a commercial software or hardware product or to provide services. LICENSEE further agrees that the PROGRAM shall not be copied or otherwise adapted in order to circumvent the need for obtaining a license for use of the PROGRAM.
*
* 3. OWNERSHIP OF INTELLECTUAL PROPERTY
* LICENSEE acknowledges that title to the PROGRAM shall remain with BROAD. The PROGRAM is marked with the following BROAD copyright notice and notice of attribution to contributors. LICENSEE shall retain such notice on all copies. LICENSEE agrees to include appropriate attribution if any results obtained from use of the PROGRAM are included in any publication.
* Copyright 2012 Broad Institute, Inc.
* Notice of attribution: The GATK2 program was made available through the generosity of Medical and Population Genetics program at the Broad Institute, Inc.
* LICENSEE shall not use any trademark or trade name of BROAD, or any variation, adaptation, or abbreviation, of such marks or trade names, or any names of officers, faculty, students, employees, or agents of BROAD except as states above for attribution purposes.
*
* 4. INDEMNIFICATION
* LICENSEE shall indemnify, defend, and hold harmless BROAD, and their respective officers, faculty, students, employees, associated investigators and agents, and their respective successors, heirs and assigns, (Indemnitees), against any liability, damage, loss, or expense (including reasonable attorneys fees and expenses) incurred by or imposed upon any of the Indemnitees in connection with any claims, suits, actions, demands or judgments arising out of any theory of liability (including, without limitation, actions in the form of tort, warranty, or strict liability and regardless of whether such action has any factual basis) pursuant to any right or license granted under this Agreement.
*
* 5. NO REPRESENTATIONS OR WARRANTIES
* THE PROGRAM IS DELIVERED AS IS. BROAD MAKES NO REPRESENTATIONS OR WARRANTIES OF ANY KIND CONCERNING THE PROGRAM OR THE COPYRIGHT, EXPRESS OR IMPLIED, INCLUDING, WITHOUT LIMITATION, WARRANTIES OF MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE, NONINFRINGEMENT, OR THE ABSENCE OF LATENT OR OTHER DEFECTS, WHETHER OR NOT DISCOVERABLE. BROAD EXTENDS NO WARRANTIES OF ANY KIND AS TO PROGRAM CONFORMITY WITH WHATEVER USER MANUALS OR OTHER LITERATURE MAY BE ISSUED FROM TIME TO TIME.
* IN NO EVENT SHALL BROAD OR ITS RESPECTIVE DIRECTORS, OFFICERS, EMPLOYEES, AFFILIATED INVESTIGATORS AND AFFILIATES BE LIABLE FOR INCIDENTAL OR CONSEQUENTIAL DAMAGES OF ANY KIND, INCLUDING, WITHOUT LIMITATION, ECONOMIC DAMAGES OR INJURY TO PROPERTY AND LOST PROFITS, REGARDLESS OF WHETHER BROAD SHALL BE ADVISED, SHALL HAVE OTHER REASON TO KNOW, OR IN FACT SHALL KNOW OF THE POSSIBILITY OF THE FOREGOING.
*
* 6. ASSIGNMENT
* This Agreement is personal to LICENSEE and any rights or obligations assigned by LICENSEE without the prior written consent of BROAD shall be null and void.
*
* 7. MISCELLANEOUS
* 7.1 Export Control. LICENSEE gives assurance that it will comply with all United States export control laws and regulations controlling the export of the PROGRAM, including, without limitation, all Export Administration Regulations of the United States Department of Commerce. Among other things, these laws and regulations prohibit, or require a license for, the export of certain types of software to specified countries.
* 7.2 Termination. LICENSEE shall have the right to terminate this Agreement for any reason upon prior written notice to BROAD. If LICENSEE breaches any provision hereunder, and fails to cure such breach within thirty (30) days, BROAD may terminate this Agreement immediately. Upon termination, LICENSEE shall provide BROAD with written assurance that the original and all copies of the PROGRAM have been destroyed, except that, upon prior written authorization from BROAD, LICENSEE may retain a copy for archive purposes.
* 7.3 Survival. The following provisions shall survive the expiration or termination of this Agreement: Articles 1, 3, 4, 5 and Sections 2.2, 2.3, 7.3, and 7.4.
* 7.4 Notice. Any notices under this Agreement shall be in writing, shall specifically refer to this Agreement, and shall be sent by hand, recognized national overnight courier, confirmed facsimile transmission, confirmed electronic mail, or registered or certified mail, postage prepaid, return receipt requested. All notices under this Agreement shall be deemed effective upon receipt.
* 7.5 Amendment and Waiver; Entire Agreement. This Agreement may be amended, supplemented, or otherwise modified only by means of a written instrument signed by all parties. Any waiver of any rights or failure to act in a specific instance shall relate only to such instance and shall not be construed as an agreement to waive any rights or fail to act in any other instance, whether or not similar. This Agreement constitutes the entire agreement among the parties with respect to its subject matter and supersedes prior agreements or understandings between the parties relating to its subject matter.
* 7.6 Binding Effect; Headings. This Agreement shall be binding upon and inure to the benefit of the parties and their respective permitted successors and assigns. All headings are for convenience only and shall not affect the meaning of any provision of this Agreement.
* 7.7 Governing Law. This Agreement shall be construed, governed, interpreted and applied in accordance with the internal laws of the Commonwealth of Massachusetts, U.S.A., without regard to conflict of laws principles.
*/
package org.broadinstitute.sting.gatk.walkers.annotator;
import org.testng.Assert;
import org.testng.annotations.DataProvider;
import org.testng.annotations.Test;
import java.util.ArrayList;
import java.lang.Integer;
import java.util.List;
/**
* Created by haasb on 3/5/14.
*/
public class StrandOddsRatioUnitTest {
private static double DELTA_PRECISION = 0.001;
@DataProvider(name = "UsingSOR")
public Object[][] makeUsingSORData() {
List<Object[]> tests = new ArrayList<>();
tests.add(new Object[]{0, 0, 0, 0, 2.0});
tests.add(new Object[]{100000, 100000, 100000, 100000, 2.0} );
tests.add(new Object[]{Integer.MAX_VALUE, Integer.MAX_VALUE, Integer.MAX_VALUE, Integer.MAX_VALUE, 2.0} );
tests.add(new Object[]{0, 0, 100000, 100000, 2.0});
tests.add(new Object[]{0, 0, Integer.MAX_VALUE, Integer.MAX_VALUE, 2.0});
tests.add(new Object[]{100000,100000,100000,0, 1000001.000001});
tests.add(new Object[]{100,100,100,0, 1001.000999});
tests.add(new Object[]{Integer.MAX_VALUE,Integer.MAX_VALUE,Integer.MAX_VALUE,0, 21474836471.0});
tests.add(new Object[]{13736,9047,41,1433, 52.95947});
tests.add(new Object[]{66, 14, 64, 4, 3.63482});
tests.add(new Object[]{351169, 306836, 153739, 2379, 56.48043});
tests.add(new Object[]{116449, 131216, 289, 16957, 52.07302});
tests.add(new Object[]{137, 159, 9, 23, 2.64460});
tests.add(new Object[]{129, 90, 21, 20, 2.09757});
tests.add(new Object[]{14054, 9160, 16, 7827, 745.89657});
tests.add(new Object[]{32803, 9184, 32117, 3283, 3.10399});
tests.add(new Object[]{2068, 6796, 1133, 0, 37235.43791});
return tests.toArray(new Object[][]{});
}
@Test(dataProvider = "UsingSOR")
public void testUsingSOR(final int refpos, final int refneg, final int altpos, final int altneg, double expectedOddsRatio ) {
int[][] contingencyTable = new int[2][2];
contingencyTable[0][0] = refpos;
contingencyTable[0][1] = refneg;
contingencyTable[1][0] = altpos;
contingencyTable[1][1] = altneg;
final double ratio = new StrandOddsRatio().symmetricOddsRatio(contingencyTable);
Assert.assertEquals(ratio, expectedOddsRatio, DELTA_PRECISION, "Pass");
}
}