/*
* 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 415 Main Street, 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 GATK3 and related documentation, if any, as they exist on the EFFECTIVE DATE and can be downloaded from http://www.broadinstitute.org/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. LICENSEE hereby automatically grants to BROAD a non-exclusive, royalty-free, irrevocable license to any LICENSEE bug fixes or modifications to the PROGRAM with unlimited rights to sublicense and/or distribute. LICENSEE agrees to provide any such modifications and bug fixes to BROAD promptly upon their creation.
* 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. PHONE-HOME FEATURE
* LICENSEE expressly acknowledges that the PROGRAM contains an embedded automatic reporting system (“PHONE-HOME”) which is enabled by default upon download. Unless LICENSEE requests disablement of PHONE-HOME, LICENSEE agrees that BROAD may collect limited information transmitted by PHONE-HOME regarding LICENSEE and its use of the PROGRAM. Such information shall include LICENSEE’S user identification, version number of the PROGRAM and tools being run, mode of analysis employed, and any error reports generated during run-time. Collection of such information is used by BROAD solely to monitor usage rates, fulfill reporting requirements to BROAD funding agencies, drive improvements to the PROGRAM, and facilitate adjustments to PROGRAM-related documentation.
*
* 4. 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 2012-2014 Broad Institute, Inc.
* Notice of attribution: The GATK3 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.
*
* 5. 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.
*
* 6. 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.
*
* 7. 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.
*
* 8. MISCELLANEOUS
* 8.1 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.
* 8.2 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.
* 8.3 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.
* 8.4 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.
* 8.5 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.
* 8.6 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.
* 8.7 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.gatk.tools.walkers.annotator;
import org.broadinstitute.gatk.engine.GenomeAnalysisEngine;
import org.broadinstitute.gatk.engine.contexts.AlignmentContext;
import org.broadinstitute.gatk.engine.contexts.ReferenceContext;
import org.broadinstitute.gatk.engine.refdata.RefMetaDataTracker;
import org.broadinstitute.gatk.tools.walkers.annotator.interfaces.ActiveRegionBasedAnnotation;
import org.broadinstitute.gatk.tools.walkers.annotator.interfaces.AnnotatorCompatible;
import org.broadinstitute.gatk.tools.walkers.annotator.interfaces.InfoFieldAnnotation;
import org.broadinstitute.gatk.utils.genotyper.MostLikelyAllele;
import org.broadinstitute.gatk.utils.genotyper.PerReadAlleleLikelihoodMap;
import org.broadinstitute.gatk.utils.MannWhitneyU;
import org.broadinstitute.gatk.utils.QualityUtils;
import org.broadinstitute.gatk.utils.sam.GATKSAMRecord;
import htsjdk.variant.vcf.VCFHeaderLine;
import org.broadinstitute.gatk.utils.collections.Pair;
import org.broadinstitute.gatk.utils.pileup.PileupElement;
import org.broadinstitute.gatk.utils.pileup.ReadBackedPileup;
import htsjdk.variant.variantcontext.Allele;
import htsjdk.variant.variantcontext.Genotype;
import htsjdk.variant.variantcontext.GenotypesContext;
import htsjdk.variant.variantcontext.VariantContext;
import java.util.*;
/**
* Abstract root for all RankSum based annotations
*/
public abstract class RankSumTest extends InfoFieldAnnotation implements ActiveRegionBasedAnnotation {
static final boolean DEBUG = false;
private boolean useDithering = true;
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) {
// either stratifiedContexts or stratifiedPerReadAlleleLikelihoodMap has to be non-null
final GenotypesContext genotypes = vc.getGenotypes();
if (genotypes == null || genotypes.size() == 0)
return null;
final ArrayList<Double> refQuals = new ArrayList<>();
final ArrayList<Double> altQuals = new ArrayList<>();
for ( final Genotype genotype : genotypes.iterateInSampleNameOrder() ) {
boolean usePileup = true;
if ( stratifiedPerReadAlleleLikelihoodMap != null ) {
final PerReadAlleleLikelihoodMap likelihoodMap = stratifiedPerReadAlleleLikelihoodMap.get(genotype.getSampleName());
if ( likelihoodMap != null && !likelihoodMap.isEmpty() ) {
fillQualsFromLikelihoodMap(vc.getAlleles(), vc.getStart(), likelihoodMap, refQuals, altQuals);
usePileup = false;
}
}
// the old UG SNP-only path through the annotations
if ( usePileup && stratifiedContexts != null ) {
final AlignmentContext context = stratifiedContexts.get(genotype.getSampleName());
if ( context != null ) {
final ReadBackedPileup pileup = context.getBasePileup();
if ( pileup != null )
fillQualsFromPileup(vc.getAlleles(), pileup, refQuals, altQuals);
}
}
}
if ( refQuals.isEmpty() && altQuals.isEmpty() )
return null;
final MannWhitneyU mannWhitneyU = new MannWhitneyU(useDithering);
for (final Double qual : altQuals) {
mannWhitneyU.add(qual, MannWhitneyU.USet.SET1);
}
for (final Double qual : refQuals) {
mannWhitneyU.add(qual, MannWhitneyU.USet.SET2);
}
if (DEBUG) {
System.out.format("%s, REF QUALS:", this.getClass().getName());
for (final Double qual : refQuals)
System.out.format("%4.1f ", qual);
System.out.println();
System.out.format("%s, ALT QUALS:", this.getClass().getName());
for (final Double qual : altQuals)
System.out.format("%4.1f ", qual);
System.out.println();
}
// we are testing that set1 (the alt bases) have lower quality scores than set2 (the ref bases)
final Pair<Double, Double> testResults = mannWhitneyU.runOneSidedTest(MannWhitneyU.USet.SET1);
final Map<String, Object> map = new HashMap<>();
if (!Double.isNaN(testResults.first))
map.put(getKeyNames().get(0), String.format("%.3f", testResults.first));
return map;
}
private void fillQualsFromPileup(final List<Allele> alleles,
final ReadBackedPileup pileup,
final List<Double> refQuals,
final List<Double> altQuals) {
for ( final PileupElement p : pileup ) {
if ( isUsableBase(p) ) {
final Double value = getElementForPileupElement(p);
if ( value == null )
continue;
if ( alleles.get(0).equals(Allele.create(p.getBase(), true)) )
refQuals.add(value);
else if ( alleles.contains(Allele.create(p.getBase())) )
altQuals.add(value);
}
}
}
private void fillQualsFromLikelihoodMap(final List<Allele> alleles,
final int refLoc,
final PerReadAlleleLikelihoodMap likelihoodMap,
final List<Double> refQuals,
final List<Double> altQuals) {
for ( final Map.Entry<GATKSAMRecord, Map<Allele,Double>> el : likelihoodMap.getLikelihoodReadMap().entrySet() ) {
final MostLikelyAllele a = PerReadAlleleLikelihoodMap.getMostLikelyAllele(el.getValue());
if ( ! a.isInformative() )
continue; // read is non-informative
final GATKSAMRecord read = el.getKey();
if ( isUsableRead(read, refLoc) ) {
final Double value = getElementForRead(read, refLoc, a);
if ( value == null )
continue;
if ( a.getMostLikelyAllele().isReference() )
refQuals.add(value);
else if ( alleles.contains(a.getMostLikelyAllele()) )
altQuals.add(value);
}
}
}
/**
* Get the element for the given read at the given reference position
*
* @param read the read
* @param refLoc the reference position
* @param mostLikelyAllele the most likely allele for this read
* @return a Double representing the element to be used in the rank sum test, or null if it should not be used
*/
protected Double getElementForRead(final GATKSAMRecord read, final int refLoc, final MostLikelyAllele mostLikelyAllele) {
return getElementForRead(read, refLoc);
}
/**
* Get the element for the given read at the given reference position
*
* @param read the read
* @param refLoc the reference position
* @return a Double representing the element to be used in the rank sum test, or null if it should not be used
*/
protected abstract Double getElementForRead(final GATKSAMRecord read, final int refLoc);
// TODO -- until the ReadPosRankSumTest stops treating these differently, we need to have separate methods for GATKSAMRecords and PileupElements. Yuck.
/**
* Get the element for the given read at the given reference position
*
* By default this function returns null, indicating that the test doesn't support the old style of pileup calculations
*
* @param p the pileup element
* @return a Double representing the element to be used in the rank sum test, or null if it should not be used
*/
protected Double getElementForPileupElement(final PileupElement p) {
// does not work in pileup mode
return null;
}
/**
* Can the base in this pileup element be used in comparative tests between ref / alt bases?
*
* Note that this function by default does not allow deletion pileup elements
*
* @param p the pileup element to consider
* @return true if this base is part of a meaningful read for comparison, false otherwise
*/
protected boolean isUsableBase(final PileupElement p) {
return !(p.isDeletion() ||
p.getMappingQual() == 0 ||
p.getMappingQual() == QualityUtils.MAPPING_QUALITY_UNAVAILABLE ||
((int) p.getQual()) < QualityUtils.MIN_USABLE_Q_SCORE); // need the unBAQed quality score here
}
/**
* Can the read be used in comparative tests between ref / alt bases?
*
* @param read the read to consider
* @param refLoc the reference location
* @return true if this read is meaningful for comparison, false otherwise
*/
protected boolean isUsableRead(final GATKSAMRecord read, final int refLoc) {
return !( read.getMappingQuality() == 0 ||
read.getMappingQuality() == QualityUtils.MAPPING_QUALITY_UNAVAILABLE );
}
/**
* Initialize the rank sum test annotation using walker and engine information. Right now this checks to see if
* engine randomization is turned off, and if so does not dither.
* @param walker the walker
* @param toolkit the GATK engine
* @param headerLines the header lines
*/
public void initialize ( AnnotatorCompatible walker, GenomeAnalysisEngine toolkit, Set<VCFHeaderLine> headerLines ) {
useDithering = ! toolkit.getArguments().disableDithering;
}
}