+/*
+ * Jalview - A Sequence Alignment Editor and Viewer ($$Version-Rel$$)
+ * Copyright (C) $$Year-Rel$$ The Jalview Authors
+ *
+ * This file is part of Jalview.
+ *
+ * Jalview is free software: you can redistribute it and/or
+ * modify it under the terms of the GNU General Public License
+ * as published by the Free Software Foundation, either version 3
+ * of the License, or (at your option) any later version.
+ *
+ * Jalview is distributed in the hope that it will be useful, but
+ * WITHOUT ANY WARRANTY; without even the implied warranty
+ * of MERCHANTABILITY or FITNESS FOR A PARTICULAR
+ * PURPOSE. See the GNU General Public License for more details.
+ *
+ * You should have received a copy of the GNU General Public License
+ * along with Jalview. If not, see <http://www.gnu.org/licenses/>.
+ * The Jalview Authors are detailed in the 'AUTHORS' file.
+ */
package jalview.io.vcf;
-import jalview.analysis.Dna;
-import jalview.api.AlignViewControllerGuiI;
-import jalview.bin.Cache;
-import jalview.datamodel.DBRefEntry;
-import jalview.datamodel.GeneLociI;
-import jalview.datamodel.Mapping;
-import jalview.datamodel.SequenceFeature;
-import jalview.datamodel.SequenceI;
-import jalview.datamodel.features.FeatureAttributeType;
-import jalview.datamodel.features.FeatureSource;
-import jalview.datamodel.features.FeatureSources;
-import jalview.ext.ensembl.EnsemblMap;
-import jalview.ext.htsjdk.HtsContigDb;
-import jalview.ext.htsjdk.VCFReader;
-import jalview.io.gff.Gff3Helper;
-import jalview.io.gff.SequenceOntologyI;
-import jalview.util.MapList;
-import jalview.util.MappingUtils;
-import jalview.util.MessageManager;
+import java.util.Locale;
import java.io.File;
import java.io.IOException;
-import java.io.UnsupportedEncodingException;
-import java.net.URLDecoder;
import java.util.ArrayList;
import java.util.HashMap;
import java.util.HashSet;
import htsjdk.variant.vcf.VCFHeaderLineCount;
import htsjdk.variant.vcf.VCFHeaderLineType;
import htsjdk.variant.vcf.VCFInfoHeaderLine;
+import jalview.analysis.Dna;
+import jalview.api.AlignViewControllerGuiI;
+import jalview.bin.Cache;
+import jalview.datamodel.DBRefEntry;
+import jalview.datamodel.GeneLociI;
+import jalview.datamodel.Mapping;
+import jalview.datamodel.SequenceFeature;
+import jalview.datamodel.SequenceI;
+import jalview.datamodel.features.FeatureAttributeType;
+import jalview.datamodel.features.FeatureSource;
+import jalview.datamodel.features.FeatureSources;
+import jalview.ext.ensembl.EnsemblMap;
+import jalview.ext.htsjdk.HtsContigDb;
+import jalview.ext.htsjdk.VCFReader;
+import jalview.io.gff.Gff3Helper;
+import jalview.io.gff.SequenceOntologyI;
+import jalview.util.MapList;
+import jalview.util.MappingUtils;
+import jalview.util.MessageManager;
+import jalview.util.StringUtils;
/**
* A class to read VCF data (using the htsjdk) and add variants as sequence
*/
public class VCFLoader
{
- private static final String ENCODED_COMMA = "%2C";
-
- private static final String ENCODED_PERCENT = "%25";
-
- private static final String ENCODED_EQUALS = "%3D";
-
- private static final String ENCODED_SEMICOLON = "%3B";
-
- private static final String ENCODED_COLON = "%3A";
-
- private static final String UTF_8 = "UTF-8";
+ private static final String VCF_ENCODABLE = ":;=%,";
/*
* Jalview feature attributes for VCF fixed column data
VCFHeaderLine headerLine = header.getOtherHeaderLine(VCFHeader.REFERENCE_KEY);
if (headerLine == null)
{
- Cache.log.error("VCF reference header not found");
+ Cache.error("VCF reference header not found");
return null;
}
String ref = headerLine.getValue();
}
else
{
- Cache.log.error("VCF reference not found: " + ref);
+ Cache.error("VCF reference not found: " + ref);
}
return seq;
{
if (reference == null)
{
- Cache.log.error("No VCF ##reference found, defaulting to "
+ Cache.error("No VCF ##reference found, defaulting to "
+ DEFAULT_REFERENCE + ":" + DEFAULT_SPECIES);
reference = DEFAULT_REFERENCE; // default to GRCh37 if not specified
}
- reference = reference.toLowerCase();
+ reference = reference.toLowerCase(Locale.ROOT);
/*
* for a non-human species, or other assembly identifier,
String[] tokens = token.split("=");
if (tokens.length == 2)
{
- if (reference.contains(tokens[0].trim().toLowerCase()))
+ if (reference.contains(tokens[0].trim().toLowerCase(Locale.ROOT)))
{
vcfAssembly = tokens[1].trim();
break;
String[] tokens = token.split("=");
if (tokens.length == 2)
{
- if (reference.contains(tokens[0].trim().toLowerCase()))
+ if (reference.contains(tokens[0].trim().toLowerCase(Locale.ROOT)))
{
vcfSpecies = tokens[1].trim();
break;
{
for (Pattern p : filters)
{
- if (p.matcher(id.toUpperCase()).matches())
+ if (p.matcher(id.toUpperCase(Locale.ROOT)).matches())
{
return true;
}
{
try
{
- patterns.add(Pattern.compile(token.toUpperCase()));
+ patterns.add(Pattern.compile(token.toUpperCase(Locale.ROOT)));
} catch (PatternSyntaxException e)
{
System.err.println("Invalid pattern ignored: " + token);
/**
* Transfers VCF features to sequences to which this sequence has a mapping.
- * If the mapping is 3:1, computes peptide variants from nucleotide variants.
*
* @param seq
*/
GeneLociI seqCoords = seq.getGeneLoci();
if (seqCoords == null)
{
- Cache.log.warn(String.format(
+ Cache.warn(String.format(
"Can't query VCF for %s as chromosome coordinates not known",
seq.getName()));
return null;
// returned with the Ensembl sequence; todo: support aliases?
if (!vcfSpecies.equalsIgnoreCase(species))
{
- Cache.log.warn("No VCF loaded to " + seq.getName()
+ Cache.warn("No VCF loaded to " + seq.getName()
+ " as species not matched");
return null;
}
vcfAssembly);
if (newRange == null)
{
- Cache.log.error(
+ Cache.error(
String.format("Failed to map %s:%s:%s:%d:%d to %s", species,
chromosome, seqRef, range[0], range[1],
vcfAssembly));
int[] featureRange = map.map.locateInFrom(variant.getStart(),
variant.getEnd());
+ /*
+ * only take features whose range is fully mappable to sequence positions
+ */
if (featureRange != null)
{
int featureStart = Math.min(featureRange[0], featureRange[1]);
int featureEnd = Math.max(featureRange[0], featureRange[1]);
- count += addAlleleFeatures(seq, variant, featureStart,
- featureEnd, forwardStrand);
+ if (featureEnd - featureStart == variant.getEnd()
+ - variant.getStart())
+ {
+ count += addAlleleFeatures(seq, variant, featureStart,
+ featureEnd, forwardStrand);
+ }
}
}
variants.close();
* RuntimeException throwable by htsjdk
*/
String msg = String.format("Error reading VCF for %s:%d-%d: %s ",
- map.chromosome, vcfStart, vcfEnd);
- Cache.log.error(msg);
+ map.chromosome, vcfStart, vcfEnd,e.getLocalizedMessage());
+ Cache.error(msg);
}
}
*/
String consequence = getConsequenceForAlleleAndFeature(variant, CSQ_FIELD,
altAlleleIndex, csqAlleleFieldIndex,
- csqAlleleNumberFieldIndex, seq.getName().toLowerCase(),
+ csqAlleleNumberFieldIndex, seq.getName().toLowerCase(Locale.ROOT),
csqFeatureFieldIndex);
/*
{
String featureIdentifier = csqFields[featureFieldIndex];
if (featureIdentifier.length() > 4
- && seqName.indexOf(featureIdentifier.toLowerCase()) > -1)
+ && seqName.indexOf(featureIdentifier.toLowerCase(Locale.ROOT)) > -1)
{
/*
* feature (transcript) matched - now check for allele match
String value = getAttributeValue(variant, key, index);
if (value != null && isValid(variant, key, value))
{
- value = decodeSpecialCharacters(value);
+ /*
+ * decode colon, semicolon, equals sign, percent sign, comma (only)
+ * as required by the VCF specification (para 1.2)
+ */
+ value = StringUtils.urlDecode(value, VCF_ENCODABLE);
addFeatureAttribute(sf, key, value);
}
}
}
/**
- * Decodes colon, semicolon, equals sign, percent sign, comma to their decoded
- * form. The VCF specification (para 1.2) requires these to be encoded where not
- * used with their special meaning in the VCF syntax. Note that general URL
- * decoding should not be applied, since this would incorrectly decode (for
- * example) a '+' sign.
- *
- * @param value
- * @return
- */
- protected static String decodeSpecialCharacters(String value)
- {
- /*
- * avoid regex compilation if it is not needed!
- */
- if (!value.contains(ENCODED_COLON) && !value.contains(ENCODED_SEMICOLON)
- && !value.contains(ENCODED_EQUALS)
- && !value.contains(ENCODED_PERCENT)
- && !value.contains(ENCODED_COMMA))
- {
- return value;
- }
-
- value = value.replace(ENCODED_COLON, ":")
- .replace(ENCODED_SEMICOLON, ";").replace(ENCODED_EQUALS, "=")
- .replace(ENCODED_PERCENT, "%").replace(ENCODED_COMMA, ",");
- return value;
- }
-
- /**
* Answers true for '.', null, or an empty value, or if the INFO type is String.
* If the INFO type is Integer or Float, answers false if the value is not in
* valid format.
VCFInfoHeaderLine infoHeader = header.getInfoHeaderLine(infoId);
if (infoHeader == null)
{
- Cache.log.error("Field " + infoId + " has no INFO header");
+ Cache.error("Field " + infoId + " has no INFO header");
return false;
}
VCFHeaderLineType infoType = infoHeader.getType();
if (!badData.contains(token))
{
badData.add(token);
- Cache.log.error(String.format("Invalid VCF data at %s:%d %s=%s",
+ Cache.error(String.format("Invalid VCF data at %s:%d %s=%s",
variant.getContig(), variant.getStart(), infoId, value));
}
}
* VCF spec requires encoding of special characters e.g. '='
* so decode them here before storing
*/
- try
- {
- field = URLDecoder.decode(field, UTF_8);
- } catch (UnsupportedEncodingException e)
- {
- }
+ field = StringUtils.urlDecode(field, VCF_ENCODABLE);
csqValues.put(id, field);
}
}
git://source.jalview.org / jalview.git / blobdiff