X-Git-Url: http://source.jalview.org/gitweb/?a=blobdiff_plain;f=src%2Fjalview%2Fext%2Fensembl%2FEnsemblSeqProxy.java;h=577111e9003c78ea77ece4c10da1e601cfe5d91b;hb=ac0c81576e51361be0c2afa265d006b516c43def;hp=e77051d493bf6d676de3b8d03ed936bc958f07a5;hpb=46b2bc114452c0e30463214b3e82eb0a5c98d23f;p=jalview.git
diff --git a/src/jalview/ext/ensembl/EnsemblSeqProxy.java b/src/jalview/ext/ensembl/EnsemblSeqProxy.java
index e77051d..577111e 100644
--- a/src/jalview/ext/ensembl/EnsemblSeqProxy.java
+++ b/src/jalview/ext/ensembl/EnsemblSeqProxy.java
@@ -1,6 +1,28 @@
+/*
+ * 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 .
+ * The Jalview Authors are detailed in the 'AUTHORS' file.
+ */
package jalview.ext.ensembl;
import jalview.analysis.AlignmentUtils;
+import jalview.analysis.Dna;
+import jalview.bin.Cache;
import jalview.datamodel.Alignment;
import jalview.datamodel.AlignmentI;
import jalview.datamodel.DBRefEntry;
@@ -8,16 +30,16 @@ import jalview.datamodel.DBRefSource;
import jalview.datamodel.Mapping;
import jalview.datamodel.SequenceFeature;
import jalview.datamodel.SequenceI;
+import jalview.datamodel.features.SequenceFeatures;
import jalview.exceptions.JalviewException;
import jalview.io.FastaFile;
import jalview.io.FileParse;
import jalview.io.gff.SequenceOntologyFactory;
import jalview.io.gff.SequenceOntologyI;
-import jalview.schemes.ResidueProperties;
+import jalview.util.Comparison;
import jalview.util.DBRefUtils;
+import jalview.util.IntRangeComparator;
import jalview.util.MapList;
-import jalview.util.MappingUtils;
-import jalview.util.StringUtils;
import java.io.IOException;
import java.net.MalformedURLException;
@@ -25,22 +47,19 @@ import java.net.URL;
import java.util.ArrayList;
import java.util.Arrays;
import java.util.Collections;
-import java.util.Comparator;
-import java.util.LinkedHashMap;
import java.util.List;
+import java.util.Map;
import java.util.Map.Entry;
/**
* Base class for Ensembl sequence fetchers
*
+ * @see http://rest.ensembl.org/documentation/info/sequence_id
* @author gmcarstairs
*/
public abstract class EnsemblSeqProxy extends EnsemblRestClient
{
- private static final List CROSS_REFERENCES = Arrays
- .asList(new String[] { "CCDS" });
-
- protected static final String CONSEQUENCE_TYPE = "consequence_type";
+ private static final String ALLELES = "alleles";
protected static final String PARENT = "Parent";
@@ -48,6 +67,8 @@ public abstract class EnsemblSeqProxy extends EnsemblRestClient
protected static final String NAME = "Name";
+ protected static final String DESCRIPTION = "description";
+
/*
* enum for 'type' parameter to the /sequence REST service
*/
@@ -59,7 +80,7 @@ public abstract class EnsemblSeqProxy extends EnsemblRestClient
GENOMIC("genomic"),
/**
- * type=cdna to fetch dna including UTRs
+ * type=cdna to fetch coding dna including UTRs
*/
CDNA("cdna"),
@@ -92,30 +113,19 @@ public abstract class EnsemblSeqProxy extends EnsemblRestClient
}
/**
- * A comparator to sort ranges into ascending start position order
+ * Default constructor (to use rest.ensembl.org)
*/
- private class RangeSorter implements Comparator
+ public EnsemblSeqProxy()
{
- boolean forwards;
-
- RangeSorter(boolean forward)
- {
- forwards = forward;
- }
-
- @Override
- public int compare(int[] o1, int[] o2)
- {
- return (forwards ? 1 : -1) * Integer.compare(o1[0], o2[0]);
- }
-
+ super();
}
/**
- * Constructor
+ * Constructor given the target domain to fetch data from
*/
- public EnsemblSeqProxy()
+ public EnsemblSeqProxy(String d)
{
+ super(d);
}
/**
@@ -129,8 +139,8 @@ public abstract class EnsemblSeqProxy extends EnsemblRestClient
// danger: accession separator used as a regex here, a string elsewhere
// in this case it is ok (it is just a space), but (e.g.) '\' would not be
- List allIds = Arrays.asList(query
- .split(getAccessionSeparator()));
+ List allIds = Arrays
+ .asList(query.split(getAccessionSeparator()));
AlignmentI alignment = null;
inProgress = true;
@@ -153,18 +163,16 @@ public abstract class EnsemblSeqProxy extends EnsemblRestClient
+ " chunks. Unexpected problem (" + r.getLocalizedMessage()
+ ")";
System.err.println(msg);
+ r.printStackTrace();
break;
- // if (alignment != null)
- // {
- // break; // return what we got
- // }
- // else
- // {
- // throw new JalviewException(msg, r);
- // }
}
}
+ if (alignment == null)
+ {
+ return null;
+ }
+
/*
* fetch and transfer genomic sequence features,
* fetch protein product and add as cross-reference
@@ -204,11 +212,11 @@ public abstract class EnsemblSeqProxy extends EnsemblRestClient
* get 'dummy' genomic sequence with exon, cds and variation features
*/
SequenceI genomicSequence = null;
- EnsemblFeatures gffFetcher = new EnsemblFeatures();
+ EnsemblFeatures gffFetcher = new EnsemblFeatures(getDomain());
EnsemblFeatureType[] features = getFeaturesToFetch();
AlignmentI geneFeatures = gffFetcher.getSequenceRecords(accId,
features);
- if (geneFeatures.getHeight() > 0)
+ if (geneFeatures != null && geneFeatures.getHeight() > 0)
{
genomicSequence = geneFeatures.getSequenceAt(0);
}
@@ -230,8 +238,8 @@ public abstract class EnsemblSeqProxy extends EnsemblRestClient
}
} catch (IOException e)
{
- System.err.println("Error transferring Ensembl features: "
- + e.getMessage());
+ System.err.println(
+ "Error transferring Ensembl features: " + e.getMessage());
}
}
@@ -254,10 +262,11 @@ public abstract class EnsemblSeqProxy extends EnsemblRestClient
String accId = querySeq.getName();
try
{
- AlignmentI protein = new EnsemblProtein().getSequenceRecords(accId);
+ AlignmentI protein = new EnsemblProtein(getDomain())
+ .getSequenceRecords(accId);
if (protein == null || protein.getHeight() == 0)
{
- System.out.println("Failed to retrieve protein for " + accId);
+ System.out.println("No protein product found for " + accId);
return;
}
SequenceI proteinSeq = protein.getSequenceAt(0);
@@ -268,22 +277,65 @@ public abstract class EnsemblSeqProxy extends EnsemblRestClient
proteinSeq.createDatasetSequence();
querySeq.createDatasetSequence();
- MapList mapList = mapCdsToProtein(querySeq, proteinSeq);
+ MapList mapList = AlignmentUtils.mapCdsToProtein(querySeq,
+ proteinSeq);
if (mapList != null)
{
// clunky: ensure Uniprot xref if we have one is on mapped sequence
SequenceI ds = proteinSeq.getDatasetSequence();
- ds.setSourceDBRef(proteinSeq.getSourceDBRef());
+ // TODO: Verify ensp primary ref is on proteinSeq.getDatasetSequence()
Mapping map = new Mapping(ds, mapList);
- DBRefEntry dbr = new DBRefEntry(getDbSource(), getDbVersion(),
- accId, map);
+ DBRefEntry dbr = new DBRefEntry(getDbSource(),
+ getEnsemblDataVersion(), proteinSeq.getName(), map);
querySeq.getDatasetSequence().addDBRef(dbr);
-
+ DBRefEntry[] uprots = DBRefUtils.selectRefs(ds.getDBRefs(),
+ new String[]
+ { DBRefSource.UNIPROT });
+ DBRefEntry[] upxrefs = DBRefUtils.selectRefs(querySeq.getDBRefs(),
+ new String[]
+ { DBRefSource.UNIPROT });
+ if (uprots != null)
+ {
+ for (DBRefEntry up : uprots)
+ {
+ // locate local uniprot ref and map
+ List upx = DBRefUtils.searchRefs(upxrefs,
+ up.getAccessionId());
+ DBRefEntry upxref;
+ if (upx.size() != 0)
+ {
+ upxref = upx.get(0);
+
+ if (upx.size() > 1)
+ {
+ Cache.log.warn(
+ "Implementation issue - multiple uniprot acc on product sequence.");
+ }
+ }
+ else
+ {
+ upxref = new DBRefEntry(DBRefSource.UNIPROT,
+ getEnsemblDataVersion(), up.getAccessionId());
+ }
+
+ Mapping newMap = new Mapping(ds, mapList);
+ upxref.setVersion(getEnsemblDataVersion());
+ upxref.setMap(newMap);
+ if (upx.size() == 0)
+ {
+ // add the new uniprot ref
+ querySeq.getDatasetSequence().addDBRef(upxref);
+ }
+
+ }
+ }
+
/*
- * compute peptide variants from dna variants and add as
- * sequence features on the protein sequence ta-da
+ * copy exon features to protein, compute peptide variants from dna
+ * variants and add as features on the protein sequence ta-da
*/
- computeProteinFeatures(querySeq, proteinSeq, mapList);
+ AlignmentUtils.computeProteinFeatures(querySeq, proteinSeq,
+ mapList);
}
} catch (Exception e)
{
@@ -294,8 +346,7 @@ public abstract class EnsemblSeqProxy extends EnsemblRestClient
}
/**
- * Get Uniprot and PDB xrefs from Ensembl, and attach them to the protein
- * sequence
+ * Get database xrefs from Ensembl, and attach them to the sequence
*
* @param seq
*/
@@ -306,127 +357,20 @@ public abstract class EnsemblSeqProxy extends EnsemblRestClient
seq = seq.getDatasetSequence();
}
- EnsemblXref xrefFetcher = new EnsemblXref();
- List xrefs = xrefFetcher.getCrossReferences(seq.getName(),
- getCrossReferenceDatabases());
+ EnsemblXref xrefFetcher = new EnsemblXref(getDomain(), getDbSource(),
+ getEnsemblDataVersion());
+ List xrefs = xrefFetcher.getCrossReferences(seq.getName());
for (DBRefEntry xref : xrefs)
{
seq.addDBRef(xref);
- /*
- * Save any Uniprot xref to be the reference for SIFTS mapping
- */
- if (DBRefSource.UNIPROT.equals(xref.getSource()))
- {
- seq.setSourceDBRef(xref);
- }
- }
- }
-
- /**
- * Returns a list of database names to be used when fetching cross-references.
- *
- * @return
- */
- protected List getCrossReferenceDatabases()
- {
- return CROSS_REFERENCES;
- }
-
- /**
- * Returns a mapping from dna to protein by inspecting sequence features of
- * type "CDS" on the dna.
- *
- * @param dnaSeq
- * @param proteinSeq
- * @return
- */
- protected MapList mapCdsToProtein(SequenceI dnaSeq, SequenceI proteinSeq)
- {
- List ranges = new ArrayList(50);
-
- int mappedDnaLength = getCdsRanges(dnaSeq, ranges);
-
- int proteinLength = proteinSeq.getLength();
- List proteinRange = new ArrayList();
- int proteinStart = 1;
-
- /*
- * incomplete start codon may mean X at start of peptide
- * we ignore both for mapping purposes
- */
- if (proteinSeq.getCharAt(0) == 'X')
- {
- proteinStart = 2;
- proteinLength--;
}
- proteinRange.add(new int[] { proteinStart, proteinLength });
/*
- * dna length should map to protein (or protein plus stop codon)
+ * and add a reference to itself
*/
- int codesForResidues = mappedDnaLength / 3;
- if (codesForResidues == proteinLength
- || codesForResidues == (proteinLength + 1))
- {
- return new MapList(ranges, proteinRange, 3, 1);
- }
- return null;
- }
-
- /**
- * Adds CDS ranges to the ranges list, and returns the total length mapped
- * from.
- *
- * No need to worry about reverse strand dna, here since the retrieved
- * sequence is as transcribed (reverse complement for reverse strand), i.e in
- * the same sense as the peptide.
- *
- * @param dnaSeq
- * @param ranges
- * @return
- */
- protected int getCdsRanges(SequenceI dnaSeq, List ranges)
- {
- SequenceFeature[] sfs = dnaSeq.getSequenceFeatures();
- if (sfs == null)
- {
- return 0;
- }
- int mappedDnaLength = 0;
- for (SequenceFeature sf : sfs)
- {
- /*
- * process a CDS feature (or a sub-type of CDS)
- */
- if (SequenceOntologyFactory.getInstance().isA(sf.getType(),
- SequenceOntologyI.CDS))
- {
- int phase = 0;
- try {
- phase = Integer.parseInt(sf.getPhase());
- } catch (NumberFormatException e)
- {
- // ignore
- }
- /*
- * phase > 0 on first codon means 5' incomplete - skip to the start
- * of the next codon; example ENST00000496384
- */
- int begin = sf.getBegin();
- int end = sf.getEnd();
- if (ranges.isEmpty() && phase > 0)
- {
- begin += phase;
- if (begin > end)
- {
- continue; // shouldn't happen?
- }
- }
- ranges.add(new int[] { begin, end });
- mappedDnaLength += Math.abs(end - begin) + 1;
- }
- }
- return mappedDnaLength;
+ DBRefEntry self = new DBRefEntry(getDbSource(), getEnsemblDataVersion(),
+ seq.getName());
+ seq.addDBRef(self);
}
/**
@@ -439,8 +383,8 @@ public abstract class EnsemblSeqProxy extends EnsemblRestClient
* @throws JalviewException
* @throws IOException
*/
- protected AlignmentI fetchSequences(List ids, AlignmentI alignment)
- throws JalviewException, IOException
+ protected AlignmentI fetchSequences(List ids,
+ AlignmentI alignment) throws JalviewException, IOException
{
if (!isEnsemblAvailable())
{
@@ -448,18 +392,23 @@ public abstract class EnsemblSeqProxy extends EnsemblRestClient
throw new JalviewException("ENSEMBL Rest API not available.");
}
FileParse fp = getSequenceReader(ids);
+ if (fp == null)
+ {
+ return alignment;
+ }
+
FastaFile fr = new FastaFile(fp);
if (fr.hasWarningMessage())
{
- System.out.println(String.format(
- "Warning when retrieving %d ids %s\n%s", ids.size(),
- ids.toString(), fr.getWarningMessage()));
+ System.out.println(
+ String.format("Warning when retrieving %d ids %s\n%s",
+ ids.size(), ids.toString(), fr.getWarningMessage()));
}
else if (fr.getSeqs().size() != ids.size())
{
System.out.println(String.format(
- "Only retrieved %d sequences for %d query strings", fr
- .getSeqs().size(), ids.size()));
+ "Only retrieved %d sequences for %d query strings",
+ fr.getSeqs().size(), ids.size()));
}
if (fr.getSeqs().size() == 1 && fr.getSeqs().get(0).getLength() == 0)
@@ -472,9 +421,8 @@ public abstract class EnsemblSeqProxy extends EnsemblRestClient
if (fr.getSeqs().size() > 0)
{
- AlignmentI seqal = new Alignment(
- fr.getSeqsAsArray());
- for (SequenceI sq:seqal.getSequences())
+ AlignmentI seqal = new Alignment(fr.getSeqsAsArray());
+ for (SequenceI sq : seqal.getSequences())
{
if (sq.getDescription() == null)
{
@@ -484,7 +432,9 @@ public abstract class EnsemblSeqProxy extends EnsemblRestClient
if (ids.contains(name)
|| ids.contains(name.replace("ENSP", "ENST")))
{
- DBRefUtils.parseToDbRef(sq, DBRefSource.ENSEMBL, "0", name);
+ DBRefEntry dbref = DBRefUtils.parseToDbRef(sq, getDbSource(),
+ getEnsemblDataVersion(), name);
+ sq.addDBRef(dbref);
}
}
if (alignment == null)
@@ -513,7 +463,7 @@ public abstract class EnsemblSeqProxy extends EnsemblRestClient
* multiple ids go in the POST body instead
*/
StringBuffer urlstring = new StringBuffer(128);
- urlstring.append(SEQUENCE_ID_URL);
+ urlstring.append(getDomain() + "/sequence/id");
if (ids.size() == 1)
{
urlstring.append("/").append(ids.get(0));
@@ -522,11 +472,31 @@ public abstract class EnsemblSeqProxy extends EnsemblRestClient
urlstring.append("?type=").append(getSourceEnsemblType().getType());
urlstring.append(("&Accept=text/x-fasta"));
+ Map params = getAdditionalParameters();
+ if (params != null)
+ {
+ for (Entry entry : params.entrySet())
+ {
+ urlstring.append("&").append(entry.getKey()).append("=")
+ .append(entry.getValue());
+ }
+ }
+
URL url = new URL(urlstring.toString());
return url;
}
/**
+ * Override this method to add any additional x=y URL parameters needed
+ *
+ * @return
+ */
+ protected Map getAdditionalParameters()
+ {
+ return null;
+ }
+
+ /**
* A sequence/id POST request currently allows up to 50 queries
*
* @see http://rest.ensembl.org/documentation/info/sequence_id_post
@@ -590,8 +560,10 @@ public abstract class EnsemblSeqProxy extends EnsemblRestClient
protected MapList getGenomicRangesFromFeatures(SequenceI sourceSequence,
String accId, int start)
{
- SequenceFeature[] sfs = sourceSequence.getSequenceFeatures();
- if (sfs == null)
+ // SequenceFeature[] sfs = sourceSequence.getSequenceFeatures();
+ List sfs = sourceSequence.getFeatures()
+ .getPositionalFeatures();
+ if (sfs.isEmpty())
{
return null;
}
@@ -604,7 +576,7 @@ public abstract class EnsemblSeqProxy extends EnsemblRestClient
int mappedLength = 0;
int direction = 1; // forward
boolean directionSet = false;
-
+
for (SequenceFeature sf : sfs)
{
/*
@@ -619,22 +591,22 @@ public abstract class EnsemblSeqProxy extends EnsemblRestClient
if (directionSet && strand != direction)
{
// abort - mix of forward and backward
- System.err.println("Error: forward and backward strand for "
- + accId);
- return null;
- }
- direction = strand;
- directionSet = true;
-
- /*
- * add to CDS ranges, semi-sorted forwards/backwards
- */
- if (strand < 0)
- {
- regions.add(0, new int[] { sf.getEnd(), sf.getBegin() });
- }
- else
- {
+ System.err.println(
+ "Error: forward and backward strand for " + accId);
+ return null;
+ }
+ direction = strand;
+ directionSet = true;
+
+ /*
+ * add to CDS ranges, semi-sorted forwards/backwards
+ */
+ if (strand < 0)
+ {
+ regions.add(0, new int[] { sf.getEnd(), sf.getBegin() });
+ }
+ else
+ {
regions.add(new int[] { sf.getBegin(), sf.getEnd() });
}
mappedLength += Math.abs(sf.getEnd() - sf.getBegin() + 1);
@@ -649,7 +621,7 @@ public abstract class EnsemblSeqProxy extends EnsemblRestClient
}
}
}
-
+
if (regions.isEmpty())
{
System.out.println("Failed to identify target sequence for " + accId
@@ -661,11 +633,13 @@ public abstract class EnsemblSeqProxy extends EnsemblRestClient
* a final sort is needed since Ensembl returns CDS sorted within source
* (havana / ensembl_havana)
*/
- Collections.sort(regions, new RangeSorter(direction == 1));
-
- List to = Arrays.asList(new int[] { start,
- start + mappedLength - 1 });
-
+ Collections.sort(regions, direction == 1 ? IntRangeComparator.ASCENDING
+ : IntRangeComparator.DESCENDING);
+
+ List to = Arrays
+ .asList(new int[]
+ { start, start + mappedLength - 1 });
+
return new MapList(regions, to, 1, 1);
}
@@ -701,35 +675,103 @@ public abstract class EnsemblSeqProxy extends EnsemblRestClient
* @param mapping
* mapping from the sequence feature's coordinates to the target
* sequence
+ * @param forwardStrand
*/
protected void transferFeature(SequenceFeature sf,
- SequenceI targetSequence, MapList mapping)
+ SequenceI targetSequence, MapList mapping, boolean forwardStrand)
{
int start = sf.getBegin();
int end = sf.getEnd();
int[] mappedRange = mapping.locateInTo(start, end);
-
+
if (mappedRange != null)
{
- SequenceFeature copy = new SequenceFeature(sf);
- copy.setBegin(Math.min(mappedRange[0], mappedRange[1]));
- copy.setEnd(Math.max(mappedRange[0], mappedRange[1]));
+ String group = sf.getFeatureGroup();
+ if (".".equals(group))
+ {
+ group = getDbSource();
+ }
+ int newBegin = Math.min(mappedRange[0], mappedRange[1]);
+ int newEnd = Math.max(mappedRange[0], mappedRange[1]);
+ SequenceFeature copy = new SequenceFeature(sf, newBegin, newEnd,
+ group, sf.getScore());
targetSequence.addSequenceFeature(copy);
/*
- * for sequence_variant, make an additional feature with consequence
+ * for sequence_variant on reverse strand, have to convert the allele
+ * values to their complements
*/
- if (SequenceOntologyFactory.getInstance().isA(sf.getType(),
- SequenceOntologyI.SEQUENCE_VARIANT))
+ if (!forwardStrand && SequenceOntologyFactory.getInstance()
+ .isA(sf.getType(), SequenceOntologyI.SEQUENCE_VARIANT))
{
- String consequence = (String) sf.getValue(CONSEQUENCE_TYPE);
- if (consequence != null)
- {
- SequenceFeature sf2 = new SequenceFeature("consequence",
- consequence, copy.getBegin(), copy.getEnd(), 0f,
- null);
- targetSequence.addSequenceFeature(sf2);
- }
+ reverseComplementAlleles(copy);
+ }
+ }
+ }
+
+ /**
+ * Change the 'alleles' value of a feature by converting to complementary
+ * bases, and also update the feature description to match
+ *
+ * @param sf
+ */
+ static void reverseComplementAlleles(SequenceFeature sf)
+ {
+ final String alleles = (String) sf.getValue(ALLELES);
+ if (alleles == null)
+ {
+ return;
+ }
+ StringBuilder complement = new StringBuilder(alleles.length());
+ for (String allele : alleles.split(","))
+ {
+ reverseComplementAllele(complement, allele);
+ }
+ String comp = complement.toString();
+ sf.setValue(ALLELES, comp);
+ sf.setDescription(comp);
+
+ /*
+ * replace value of "alleles=" in sf.ATTRIBUTES as well
+ * so 'output as GFF' shows reverse complement alleles
+ */
+ String atts = sf.getAttributes();
+ if (atts != null)
+ {
+ atts = atts.replace(ALLELES + "=" + alleles, ALLELES + "=" + comp);
+ sf.setAttributes(atts);
+ }
+ }
+
+ /**
+ * Makes the 'reverse complement' of the given allele and appends it to the
+ * buffer, after a comma separator if not the first
+ *
+ * @param complement
+ * @param allele
+ */
+ static void reverseComplementAllele(StringBuilder complement,
+ String allele)
+ {
+ if (complement.length() > 0)
+ {
+ complement.append(",");
+ }
+
+ /*
+ * some 'alleles' are actually descriptive terms
+ * e.g. HGMD_MUTATION, PhenCode_variation
+ * - we don't want to 'reverse complement' these
+ */
+ if (!Comparison.isNucleotideSequence(allele, true))
+ {
+ complement.append(allele);
+ }
+ else
+ {
+ for (int i = allele.length() - 1; i >= 0; i--)
+ {
+ complement.append(Dna.getComplement(allele.charAt(i)));
}
}
}
@@ -750,15 +792,23 @@ public abstract class EnsemblSeqProxy extends EnsemblRestClient
return false;
}
- SequenceFeature[] sfs = sourceSequence.getSequenceFeatures();
- MapList mapping = getGenomicRangesFromFeatures(sourceSequence, accessionId,
- targetSequence.getStart());
+// long start = System.currentTimeMillis();
+ List sfs = sourceSequence.getFeatures()
+ .getPositionalFeatures();
+ MapList mapping = getGenomicRangesFromFeatures(sourceSequence,
+ accessionId, targetSequence.getStart());
if (mapping == null)
{
return false;
}
- return transferFeatures(sfs, targetSequence, mapping, accessionId);
+ boolean result = transferFeatures(sfs, targetSequence, mapping,
+ accessionId);
+// System.out.println("transferFeatures (" + (sfs.size()) + " --> "
+// + targetSequence.getFeatures().getFeatureCount(true) + ") to "
+// + targetSequence.getName() + " took "
+// + (System.currentTimeMillis() - start) + "ms");
+ return result;
}
/**
@@ -766,37 +816,30 @@ public abstract class EnsemblSeqProxy extends EnsemblRestClient
* converted using the mapping. Features which do not overlap are ignored.
* Features whose parent is not the specified identifier are also ignored.
*
- * @param features
+ * @param sfs
* @param targetSequence
* @param mapping
* @param parentId
* @return
*/
- protected boolean transferFeatures(SequenceFeature[] features,
+ protected boolean transferFeatures(List sfs,
SequenceI targetSequence, MapList mapping, String parentId)
{
final boolean forwardStrand = mapping.isFromForwardStrand();
/*
- * sort features by start position (descending if reverse strand)
- * before transferring (in forwards order) to the target sequence
+ * sort features by start position (which corresponds to end
+ * position descending if reverse strand) so as to add them in
+ * 'forwards' order to the target sequence
*/
- Arrays.sort(features, new Comparator()
- {
- @Override
- public int compare(SequenceFeature o1, SequenceFeature o2)
- {
- int c = Integer.compare(o1.getBegin(), o2.getBegin());
- return forwardStrand ? c : -c;
- }
- });
+ SequenceFeatures.sortFeatures(sfs, forwardStrand);
boolean transferred = false;
- for (SequenceFeature sf : features)
+ for (SequenceFeature sf : sfs)
{
if (retainFeature(sf, parentId))
{
- transferFeature(sf, targetSequence, mapping);
+ transferFeature(sf, targetSequence, mapping, forwardStrand);
transferred = true;
}
}
@@ -845,267 +888,30 @@ public abstract class EnsemblSeqProxy extends EnsemblRestClient
/**
* Returns a (possibly empty) list of features on the sequence which have the
- * specified sequence ontology type (or a sub-type of it), and the given
+ * specified sequence ontology term (or a sub-type of it), and the given
* identifier as parent
*
* @param sequence
- * @param type
+ * @param term
* @param parentId
* @return
*/
protected List findFeatures(SequenceI sequence,
- String type, String parentId)
+ String term, String parentId)
{
List result = new ArrayList();
-
- SequenceFeature[] sfs = sequence.getSequenceFeatures();
- if (sfs != null) {
- SequenceOntologyI so = SequenceOntologyFactory.getInstance();
- for (SequenceFeature sf :sfs) {
- if (so.isA(sf.getType(), type))
- {
- String parent = (String) sf.getValue(PARENT);
- if (parent.equals(parentId))
- {
- result.add(sf);
- }
- }
- }
- }
- return result;
- }
-
- /**
- * Maps exon features from dna to protein, and computes variants in peptide
- * product generated by variants in dna, and adds them as sequence_variant
- * features on the protein sequence. Returns the number of variant features
- * added.
- *
- * @param dnaSeq
- * @param peptide
- * @param dnaToProtein
- */
- static int computeProteinFeatures(SequenceI dnaSeq,
- SequenceI peptide, MapList dnaToProtein)
- {
- while (dnaSeq.getDatasetSequence() != null)
- {
- dnaSeq = dnaSeq.getDatasetSequence();
- }
- while (peptide.getDatasetSequence() != null)
- {
- peptide = peptide.getDatasetSequence();
- }
-
- AlignmentUtils.transferFeatures(dnaSeq, peptide, dnaToProtein,
- SequenceOntologyI.EXON);
-
- LinkedHashMap variants = buildDnaVariantsMap(
- dnaSeq, dnaToProtein);
-
- /*
- * scan codon variations, compute peptide variants and add to peptide sequence
- */
- int count = 0;
- for (Entry variant : variants.entrySet())
- {
- int peptidePos = variant.getKey();
- String[][] codonVariants = variant.getValue();
- String residue = String.valueOf(peptide.getCharAt(peptidePos - 1)); // 0-based
- List peptideVariants = computePeptideVariants(codonVariants,
- residue);
- if (!peptideVariants.isEmpty())
- {
- String desc = StringUtils.listToDelimitedString(peptideVariants,
- ", ");
- SequenceFeature sf = new SequenceFeature(
- SequenceOntologyI.SEQUENCE_VARIANT, desc, peptidePos,
- peptidePos, 0f, null);
- peptide.addSequenceFeature(sf);
- count++;
- }
- }
- /*
- * ugly sort to get sequence features in start position order
- * - would be better to store in Sequence as a TreeSet instead?
- */
- Arrays.sort(peptide.getSequenceFeatures(),
- new Comparator()
- {
- @Override
- public int compare(SequenceFeature o1, SequenceFeature o2)
- {
- int c = Integer.compare(o1.getBegin(), o2.getBegin());
- return c == 0 ? Integer.compare(o1.getEnd(), o2.getEnd())
- : c;
- }
- });
- return count;
- }
-
- /**
- * Builds a map whose key is position in the protein sequence, and value is an
- * array of all variants for the coding codon positions
- *
- * @param dnaSeq
- * @param dnaToProtein
- * @return
- */
- static LinkedHashMap buildDnaVariantsMap(
- SequenceI dnaSeq, MapList dnaToProtein)
- {
- /*
- * map from peptide position to all variant features of the codon for it
- * LinkedHashMap ensures we add the peptide features in sequence order
- */
- LinkedHashMap variants = new LinkedHashMap();
- SequenceOntologyI so = SequenceOntologyFactory.getInstance();
-
- SequenceFeature[] dnaFeatures = dnaSeq.getSequenceFeatures();
- if (dnaFeatures == null)
- {
- return variants;
- }
-
- int dnaStart = dnaSeq.getStart();
- int[] lastCodon = null;
- int lastPeptidePostion = 0;
-
- /*
- * build a map of codon variations for peptides
- */
- for (SequenceFeature sf : dnaFeatures)
- {
- int dnaCol = sf.getBegin();
- if (dnaCol != sf.getEnd())
- {
- // not handling multi-locus variant features
- continue;
- }
- if (so.isA(sf.getType(), SequenceOntologyI.SEQUENCE_VARIANT))
- {
- int[] mapsTo = dnaToProtein.locateInTo(dnaCol, dnaCol);
- if (mapsTo == null)
- {
- // feature doesn't lie within coding region
- continue;
- }
- int peptidePosition = mapsTo[0];
- String[][] codonVariants = variants.get(peptidePosition);
- if (codonVariants == null)
- {
- codonVariants = new String[3][];
- variants.put(peptidePosition, codonVariants);
- }
-
- /*
- * extract dna variants to a string array
- */
- String alls = (String) sf.getValue("alleles");
- if (alls == null)
- {
- continue;
- }
- String[] alleles = alls.split(",");
-
- /*
- * get this peptides codon positions e.g. [3, 4, 5] or [4, 7, 10]
- */
- int[] codon = peptidePosition == lastPeptidePostion ? lastCodon
- : MappingUtils.flattenRanges(dnaToProtein.locateInFrom(
- peptidePosition, peptidePosition));
- lastPeptidePostion = peptidePosition;
- lastCodon = codon;
-
- /*
- * save nucleotide (and this variant) for each codon position
- */
- for (int codonPos = 0; codonPos < 3; codonPos++)
- {
- String nucleotide = String.valueOf(dnaSeq
- .getCharAt(codon[codonPos] - dnaStart));
- if (codon[codonPos] == dnaCol)
- {
- /*
- * record current dna base and its alleles
- */
- String[] dnaVariants = new String[alleles.length + 1];
- dnaVariants[0] = nucleotide;
- System.arraycopy(alleles, 0, dnaVariants, 1, alleles.length);
- codonVariants[codonPos] = dnaVariants;
- }
- else if (codonVariants[codonPos] == null)
- {
- /*
- * record current dna base only
- * (at least until we find any variation and overwrite it)
- */
- codonVariants[codonPos] = new String[] { nucleotide };
- }
- }
- }
- }
- return variants;
- }
-
- /**
- * Returns a sorted, non-redundant list of all peptide translations generated
- * by the given dna variants, excluding the current residue value
- *
- * @param codonVariants
- * an array of base values (acgtACGT) for codon positions 1, 2, 3
- * @param residue
- * the current residue translation
- * @return
- */
- static List computePeptideVariants(
- String[][] codonVariants, String residue)
- {
- List result = new ArrayList();
- for (String base1 : codonVariants[0])
+ List sfs = sequence.getFeatures()
+ .getFeaturesByOntology(term);
+ for (SequenceFeature sf : sfs)
{
- for (String base2 : codonVariants[1])
+ String parent = (String) sf.getValue(PARENT);
+ if (parent != null && parent.equals(parentId))
{
- for (String base3 : codonVariants[2])
- {
- String codon = base1 + base2 + base3;
- // TODO: report frameshift/insertion/deletion
- // and multiple-base variants?!
- String peptide = codon.contains("-") ? "-" : ResidueProperties
- .codonTranslate(codon);
- if (peptide != null && !result.contains(peptide)
- && !peptide.equalsIgnoreCase(residue))
- {
- result.add(peptide);
- }
- }
+ result.add(sf);
}
}
- /*
- * sort alphabetically with STOP at the end
- */
- Collections.sort(result, new Comparator()
- {
-
- @Override
- public int compare(String o1, String o2)
- {
- if ("STOP".equals(o1))
- {
- return 1;
- }
- else if ("STOP".equals(o2))
- {
- return -1;
- }
- else
- {
- return o1.compareTo(o2);
- }
- }
- });
return result;
}