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

Adding feature for Khalid: ability to exclude particular samples.

This commit is contained in:
Eric Banks 2011-08-17 22:28:22 -04:00
parent fa1db3913b
commit a7b70e6bb4
1 changed files with 29 additions and 10 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

31
public/java/src/org/broadinstitute/sting/gatk/walkers/variantutils/SelectVariants.java
View File

@ -191,7 +191,7 @@ public class SelectVariants extends RodWalker<Integer, Integer> {
protected VCFWriter vcfWriter = null; protected VCFWriter vcfWriter = null;
@Argument(fullName="sample_name", shortName="sn", doc="Include genotypes from this sample. Can be specified multiple times", required=false) @Argument(fullName="sample_name", shortName="sn", doc="Include genotypes from this sample. Can be specified multiple times", required=false)
public Set<String> sampleNames; public Set<String> sampleNames = new HashSet<String>(0);
@Argument(fullName="sample_expressions", shortName="se", doc="Regular expression to select many samples from the ROD tracks provided. Can be specified multiple times", required=false) @Argument(fullName="sample_expressions", shortName="se", doc="Regular expression to select many samples from the ROD tracks provided. Can be specified multiple times", required=false)
public Set<String> sampleExpressions ; public Set<String> sampleExpressions ;
@ -200,7 +200,19 @@ public class SelectVariants extends RodWalker<Integer, Integer> {
public Set<File> sampleFiles; public Set<File> sampleFiles;
/** /**
* Note that thse expressions are evaluated *after* the specified samples are extracted and the INFO field annotations are updated. * Note that sample exclusion takes precedence over inclusion, so that if a sample is in both lists it will be excluded.
*/
@Argument(fullName="exclude_sample_name", shortName="xl_sn", doc="Exclude genotypes from this sample. Can be specified multiple times", required=false)
public Set<String> XLsampleNames = new HashSet<String>(0);
/**
* Note that sample exclusion takes precedence over inclusion, so that if a sample is in both lists it will be excluded.
*/
@Argument(fullName="exclude_sample_file", shortName="xl_sf", doc="File containing a list of samples (one per line) to exclude. Can be specified multiple times", required=false)
public Set<File> XLsampleFiles;
/**
* Note that these expressions are evaluated *after* the specified samples are extracted and the INFO field annotations are updated.
*/ */
@Argument(shortName="select", doc="One or more criteria to use when selecting the data", required=false) @Argument(shortName="select", doc="One or more criteria to use when selecting the data", required=false)
public ArrayList<String> SELECT_EXPRESSIONS = new ArrayList<String>(); public ArrayList<String> SELECT_EXPRESSIONS = new ArrayList<String>();
@ -304,7 +316,6 @@ public class SelectVariants extends RodWalker<Integer, Integer> {
private ArrayList<Double> afBoosts = null; private ArrayList<Double> afBoosts = null;
double bkDelta = 0.0; double bkDelta = 0.0;
private PrintStream outMVFileStream = null; private PrintStream outMVFileStream = null;
@ -321,19 +332,27 @@ public class SelectVariants extends RodWalker<Integer, Integer> {
Collection<String> samplesFromFile = SampleUtils.getSamplesFromFiles(sampleFiles); Collection<String> samplesFromFile = SampleUtils.getSamplesFromFiles(sampleFiles);
Collection<String> samplesFromExpressions = SampleUtils.matchSamplesExpressions(vcfSamples, sampleExpressions); Collection<String> samplesFromExpressions = SampleUtils.matchSamplesExpressions(vcfSamples, sampleExpressions);
// first, add any requested samples
samples.addAll(samplesFromFile); samples.addAll(samplesFromFile);
samples.addAll(samplesFromExpressions); samples.addAll(samplesFromExpressions);
if (sampleNames != null)
samples.addAll(sampleNames); samples.addAll(sampleNames);
// if none were requested, we want all of them
if ( samples.isEmpty() ) { if ( samples.isEmpty() ) {
samples.addAll(vcfSamples); samples.addAll(vcfSamples);
NO_SAMPLES_SPECIFIED = true; NO_SAMPLES_SPECIFIED = true;
} }
for (String sample : samples) { // now, exclude any requested samples
Collection<String> XLsamplesFromFile = SampleUtils.getSamplesFromFiles(XLsampleFiles);
samples.removeAll(XLsamplesFromFile);
samples.removeAll(XLsampleNames);
if ( samples.size() == 0 )
throw new UserException("All samples requested to be included were also requested to be excluded.");
for ( String sample : samples )
logger.info("Including sample '" + sample + "'"); logger.info("Including sample '" + sample + "'");
}
// Initialize VCF header // Initialize VCF header
Set<VCFHeaderLine> headerLines = VCFUtils.smartMergeHeaders(vcfRods.values(), logger); Set<VCFHeaderLine> headerLines = VCFUtils.smartMergeHeaders(vcfRods.values(), logger);