*/
package jalview.analysis;
-import static jalview.io.gff.GffConstants.CLINICAL_SIGNIFICANCE;
+import java.util.ArrayList;
+import java.util.Arrays;
+import java.util.Collection;
+import java.util.Collections;
+import java.util.HashMap;
+import java.util.HashSet;
+import java.util.Iterator;
+import java.util.LinkedHashMap;
+import java.util.List;
+import java.util.Map;
+import java.util.Map.Entry;
+import java.util.NoSuchElementException;
+import java.util.Set;
+import java.util.SortedMap;
+import java.util.TreeMap;
+import jalview.bin.Cache;
+import jalview.commands.RemoveGapColCommand;
import jalview.datamodel.AlignedCodon;
import jalview.datamodel.AlignedCodonFrame;
import jalview.datamodel.AlignedCodonFrame.SequenceToSequenceMapping;
import jalview.datamodel.SequenceGroup;
import jalview.datamodel.SequenceI;
import jalview.datamodel.features.SequenceFeatures;
-import jalview.io.gff.Gff3Helper;
import jalview.io.gff.SequenceOntologyI;
import jalview.schemes.ResidueProperties;
import jalview.util.Comparison;
import jalview.util.IntRangeComparator;
import jalview.util.MapList;
import jalview.util.MappingUtils;
-import jalview.util.StringUtils;
-
-import java.io.UnsupportedEncodingException;
-import java.net.URLEncoder;
-import java.util.ArrayList;
-import java.util.Arrays;
-import java.util.Collection;
-import java.util.Collections;
-import java.util.HashMap;
-import java.util.HashSet;
-import java.util.Iterator;
-import java.util.LinkedHashMap;
-import java.util.List;
-import java.util.Map;
-import java.util.Map.Entry;
-import java.util.NoSuchElementException;
-import java.util.Set;
-import java.util.SortedMap;
-import java.util.TreeMap;
/**
* grab bag of useful alignment manipulation operations Expect these to be
*/
public class AlignmentUtils
{
-
private static final int CODON_LENGTH = 3;
private static final String SEQUENCE_VARIANT = "sequence_variant:";
- private static final String ID = "ID";
+ /*
+ * the 'id' attribute is provided for variant features fetched from
+ * Ensembl using its REST service with JSON format
+ */
+ public static final String VARIANT_ID = "id";
/**
* A data model to hold the 'normal' base value at a position, and an optional
*/
public static AlignmentI expandContext(AlignmentI core, int flankSize)
{
- List<SequenceI> sq = new ArrayList<SequenceI>();
+ List<SequenceI> sq = new ArrayList<>();
int maxoffset = 0;
for (SequenceI s : core.getSequences())
{
public static Map<String, List<SequenceI>> getSequencesByName(
AlignmentI al)
{
- Map<String, List<SequenceI>> theMap = new LinkedHashMap<String, List<SequenceI>>();
+ Map<String, List<SequenceI>> theMap = new LinkedHashMap<>();
for (SequenceI seq : al.getSequences())
{
String name = seq.getName();
List<SequenceI> seqs = theMap.get(name);
if (seqs == null)
{
- seqs = new ArrayList<SequenceI>();
+ seqs = new ArrayList<>();
theMap.put(name, seqs);
}
seqs.add(seq);
return false;
}
- Set<SequenceI> mappedDna = new HashSet<SequenceI>();
- Set<SequenceI> mappedProtein = new HashSet<SequenceI>();
+ Set<SequenceI> mappedDna = new HashSet<>();
+ Set<SequenceI> mappedProtein = new HashSet<>();
/*
* First pass - map sequences where cross-references exist. This include
{
String lastCodon = String.valueOf(cdnaSeqChars,
cdnaLength - CODON_LENGTH, CODON_LENGTH).toUpperCase();
- for (String stop : ResidueProperties.STOP)
+ for (String stop : ResidueProperties.STOP_CODONS)
{
if (lastCodon.equals(stop))
{
* allow * in protein to match untranslatable in dna
*/
final char aaRes = aaSeqChars[aaPos];
- if ((translated == null || "STOP".equals(translated)) && aaRes == '*')
+ if ((translated == null || ResidueProperties.STOP.equals(translated))
+ && aaRes == '*')
{
continue;
}
if (dnaPos == cdnaSeqChars.length - CODON_LENGTH)
{
String codon = String.valueOf(cdnaSeqChars, dnaPos, CODON_LENGTH);
- if ("STOP".equals(ResidueProperties.codonTranslate(codon)))
+ if (ResidueProperties.STOP
+ .equals(ResidueProperties.codonTranslate(codon)))
{
return true;
}
System.err.println("Wrong alignment type in alignProteinAsDna");
return 0;
}
- List<SequenceI> unmappedProtein = new ArrayList<SequenceI>();
+ List<SequenceI> unmappedProtein = new ArrayList<>();
Map<AlignedCodon, Map<SequenceI, AlignedCodon>> alignedCodons = buildCodonColumnsMap(
protein, dna, unmappedProtein);
return alignProteinAs(protein, alignedCodons, unmappedProtein);
.findMappingsForSequence(cdsSeq, mappings);
for (AlignedCodonFrame mapping : dnaMappings)
{
- SequenceI peptide = mapping.findAlignedSequence(cdsSeq, protein);
+ List<SequenceToSequenceMapping> foundMap=new ArrayList<>();
+ SequenceI peptide = mapping.findAlignedSequence(cdsSeq, protein,foundMap);
if (peptide != null)
{
final int peptideLength = peptide.getLength();
- Mapping map = mapping.getMappingBetween(cdsSeq, peptide);
+ Mapping map = foundMap.get(0).getMapping();
if (map != null)
{
MapList mapList = map.getMap();
* {dnaSequence, {proteinSequence, codonProduct}} at that position. The
* comparator keeps the codon positions ordered.
*/
- Map<AlignedCodon, Map<SequenceI, AlignedCodon>> alignedCodons = new TreeMap<AlignedCodon, Map<SequenceI, AlignedCodon>>(
+ Map<AlignedCodon, Map<SequenceI, AlignedCodon>> alignedCodons = new TreeMap<>(
new CodonComparator());
for (SequenceI dnaSeq : dna.getSequences())
// TODO delete this ugly hack once JAL-2022 is resolved
// i.e. we can model startPhase > 0 (incomplete start codon)
- List<SequenceI> sequencesChecked = new ArrayList<SequenceI>();
+ List<SequenceI> sequencesChecked = new ArrayList<>();
AlignedCodon lastCodon = null;
- Map<SequenceI, AlignedCodon> toAdd = new HashMap<SequenceI, AlignedCodon>();
+ Map<SequenceI, AlignedCodon> toAdd = new HashMap<>();
for (Entry<AlignedCodon, Map<SequenceI, AlignedCodon>> entry : alignedCodons
.entrySet())
Map<SequenceI, AlignedCodon> seqProduct = alignedCodons.get(codon);
if (seqProduct == null)
{
- seqProduct = new HashMap<SequenceI, AlignedCodon>();
+ seqProduct = new HashMap<>();
alignedCodons.put(codon, seqProduct);
}
seqProduct.put(protein, codon);
{
continue;
}
- final List<AlignmentAnnotation> result = new ArrayList<AlignmentAnnotation>();
+ final List<AlignmentAnnotation> result = new ArrayList<>();
for (AlignmentAnnotation dsann : datasetAnnotations)
{
/*
throw new IllegalArgumentException(
"IMPLEMENTATION ERROR: dataset.getDataset() must be null!");
}
- List<SequenceI> foundSeqs = new ArrayList<SequenceI>();
- List<SequenceI> cdsSeqs = new ArrayList<SequenceI>();
+ List<SequenceI> foundSeqs = new ArrayList<>();
+ List<SequenceI> cdsSeqs = new ArrayList<>();
List<AlignedCodonFrame> mappings = dataset.getCodonFrames();
HashSet<SequenceI> productSeqs = null;
if (products != null)
{
- productSeqs = new HashSet<SequenceI>();
+ productSeqs = new HashSet<>();
for (SequenceI seq : products)
{
productSeqs.add(seq.getDatasetSequence() == null ? seq : seq
cdsSeqs.add(cdsSeq);
- if (!dataset.getSequences().contains(cdsSeqDss))
- {
- // check if this sequence is a newly created one
- // so needs adding to the dataset
- dataset.addSequence(cdsSeqDss);
- }
-
/*
- * add a mapping from CDS to the (unchanged) mapped to range
+ * build the mapping from CDS to protein
*/
- List<int[]> cdsRange = Collections.singletonList(new int[] { 1,
- cdsSeq.getLength() });
+ List<int[]> cdsRange = Collections
+ .singletonList(new int[]
+ { cdsSeq.getStart(),
+ cdsSeq.getLength() + cdsSeq.getStart() - 1 });
MapList cdsToProteinMap = new MapList(cdsRange,
mapList.getToRanges(), mapList.getFromRatio(),
mapList.getToRatio());
- AlignedCodonFrame cdsToProteinMapping = new AlignedCodonFrame();
- cdsToProteinMapping.addMap(cdsSeqDss, proteinProduct,
- cdsToProteinMap);
- /*
- * guard against duplicating the mapping if repeating this action
- */
- if (!mappings.contains(cdsToProteinMapping))
+ if (!dataset.getSequences().contains(cdsSeqDss))
{
- mappings.add(cdsToProteinMapping);
+ /*
+ * if this sequence is a newly created one, add it to the dataset
+ * and made a CDS to protein mapping (if sequence already exists,
+ * CDS-to-protein mapping _is_ the transcript-to-protein mapping)
+ */
+ dataset.addSequence(cdsSeqDss);
+ AlignedCodonFrame cdsToProteinMapping = new AlignedCodonFrame();
+ cdsToProteinMapping.addMap(cdsSeqDss, proteinProduct,
+ cdsToProteinMap);
+
+ /*
+ * guard against duplicating the mapping if repeating this action
+ */
+ if (!mappings.contains(cdsToProteinMapping))
+ {
+ mappings.add(cdsToProteinMapping);
+ }
}
propagateDBRefsToCDS(cdsSeqDss, dnaSeq.getDatasetSequence(),
return;
}
- MapList newMap = targetToFrom.traverse(fromLoci.getMap());
+ MapList newMap = targetToFrom.traverse(fromLoci.getMapping());
if (newMap != null)
{
* @param seqMappings
* the set of mappings involving dnaSeq
* @param aMapping
- * an initial candidate from seqMappings
+ * a transcript-to-peptide mapping
* @return
*/
static SequenceI findCdsForProtein(List<AlignedCodonFrame> mappings,
if (mappedFromLength == dnaLength
|| mappedFromLength == dnaLength - CODON_LENGTH)
{
- return seqDss;
+ /*
+ * if sequence has CDS features, this is a transcript with no UTR
+ * - do not take this as the CDS sequence! (JAL-2789)
+ */
+ if (seqDss.getFeatures().getFeaturesByOntology(SequenceOntologyI.CDS)
+ .isEmpty())
+ {
+ return seqDss;
+ }
}
/*
{
/*
* found a 3:1 mapping to the protein product which covers
- * the whole dna sequence i.e. is from CDS; finally check it
- * is from the dna start sequence
+ * the whole dna sequence i.e. is from CDS; finally check the CDS
+ * is mapped from the given dna start sequence
*/
SequenceI cdsSeq = map.getFromSeq();
+ // todo this test is weak if seqMappings contains multiple mappings;
+ // we get away with it if transcript:cds relationship is 1:1
List<AlignedCodonFrame> dnaToCdsMaps = MappingUtils
.findMappingsForSequence(cdsSeq, seqMappings);
if (!dnaToCdsMaps.isEmpty())
static SequenceI makeCdsSequence(SequenceI seq, Mapping mapping,
AlignmentI dataset)
{
+ /*
+ * construct CDS sequence name as "CDS|" with 'from id' held in the mapping
+ * if set (e.g. EMBL protein_id), else sequence name appended
+ */
+ String mapFromId = mapping.getMappedFromId();
+ final String seqId = "CDS|"
+ + (mapFromId != null ? mapFromId : seq.getName());
+
+ SequenceI newSeq = null;
+
+ /*
+ * construct CDS sequence by splicing mapped from ranges
+ */
char[] seqChars = seq.getSequence();
List<int[]> fromRanges = mapping.getMap().getFromRanges();
int cdsWidth = MappingUtils.getLength(fromRanges);
newSeqChars[newPos++] = Dna.getComplement(seqChars[i - 1]);
}
}
+
+ newSeq = new Sequence(seqId, newSeqChars, 1, newPos);
}
- /*
- * assign 'from id' held in the mapping if set (e.g. EMBL protein_id),
- * else generate a sequence name
- */
- String mapFromId = mapping.getMappedFromId();
- String seqId = "CDS|" + (mapFromId != null ? mapFromId : seq.getName());
- SequenceI newSeq = new Sequence(seqId, newSeqChars, 1, newPos);
if (dataset != null)
{
SequenceI[] matches = dataset.findSequenceMatch(newSeq.getName());
}
else
{
- System.err.println(
- "JAL-2154 regression: warning - found (and ignnored a duplicate CDS sequence):"
- + mtch.toString());
+ Cache.log.error(
+ "JAL-2154 regression: warning - found (and ignored) a duplicate CDS sequence:" + mtch.toString());
}
}
}
protected static List<DBRefEntry> propagateDBRefsToCDS(SequenceI cdsSeq,
SequenceI contig, SequenceI proteinProduct, Mapping mapping)
{
+
// gather direct refs from contig congruent with mapping
- List<DBRefEntry> direct = new ArrayList<DBRefEntry>();
- HashSet<String> directSources = new HashSet<String>();
+ List<DBRefEntry> direct = new ArrayList<>();
+ HashSet<String> directSources = new HashSet<>();
+
if (contig.getDBRefs() != null)
{
for (DBRefEntry dbr : contig.getDBRefs())
DBRefEntry[] onSource = DBRefUtils.selectRefs(
proteinProduct.getDBRefs(),
directSources.toArray(new String[0]));
- List<DBRefEntry> propagated = new ArrayList<DBRefEntry>();
+ List<DBRefEntry> propagated = new ArrayList<>();
// and generate appropriate mappings
for (DBRefEntry cdsref : direct)
{
copyTo = copyTo.getDatasetSequence();
}
+ if (fromSeq == copyTo || fromSeq.getDatasetSequence() == copyTo)
+ {
+ return 0; // shared dataset sequence
+ }
/*
* get features, optionally restricted by an ontology term
/**
* Returns a mapping from dna to protein by inspecting sequence features of
- * type "CDS" on the dna.
+ * type "CDS" on the dna. A mapping is constructed if the total CDS feature
+ * length is 3 times the peptide length (optionally after dropping a trailing
+ * stop codon). This method does not check whether the CDS nucleotide sequence
+ * translates to the peptide sequence.
*
* @param dnaSeq
* @param proteinSeq
List<int[]> ranges = findCdsPositions(dnaSeq);
int mappedDnaLength = MappingUtils.getLength(ranges);
+ /*
+ * if not a whole number of codons, truncate mapping
+ */
+ int codonRemainder = mappedDnaLength % CODON_LENGTH;
+ if (codonRemainder > 0)
+ {
+ mappedDnaLength -= codonRemainder;
+ MappingUtils.removeEndPositions(codonRemainder, ranges);
+ }
+
int proteinLength = proteinSeq.getLength();
int proteinStart = proteinSeq.getStart();
int proteinEnd = proteinSeq.getEnd();
proteinStart++;
proteinLength--;
}
- List<int[]> proteinRange = new ArrayList<int[]>();
+ List<int[]> proteinRange = new ArrayList<>();
/*
* dna length should map to protein (or protein plus stop codon)
if (codesForResidues == (proteinLength + 1))
{
// assuming extra codon is for STOP and not in peptide
+ // todo: check trailing codon is indeed a STOP codon
codesForResidues--;
+ mappedDnaLength -= CODON_LENGTH;
+ MappingUtils.removeEndPositions(CODON_LENGTH, ranges);
}
+
if (codesForResidues == proteinLength)
{
proteinRange.add(new int[] { proteinStart, proteinEnd });
/**
* Returns a list of CDS ranges found (as sequence positions base 1), i.e. of
- * start/end positions of sequence features of type "CDS" (or a sub-type of
+ * [start, end] positions of sequence features of type "CDS" (or a sub-type of
* CDS in the Sequence Ontology). The ranges are sorted into ascending start
* position order, so this method is only valid for linear CDS in the same
* sense as the protein product.
*/
protected static List<int[]> findCdsPositions(SequenceI dnaSeq)
{
- List<int[]> result = new ArrayList<int[]>();
+ List<int[]> result = new ArrayList<>();
List<SequenceFeature> sfs = dnaSeq.getFeatures().getFeaturesByOntology(
SequenceOntologyI.CDS);
return result;
}
SequenceFeatures.sortFeatures(sfs, true);
- int startPhase = 0;
for (SequenceFeature sf : sfs)
{
*/
int begin = sf.getBegin();
int end = sf.getEnd();
- if (result.isEmpty())
+ if (result.isEmpty() && phase > 0)
{
begin += phase;
if (begin > end)
}
/*
- * remove 'startPhase' positions (usually 0) from the first range
- * so we begin at the start of a complete codon
- */
- if (!result.isEmpty())
- {
- // TODO JAL-2022 correctly model start phase > 0
- result.get(0)[0] += startPhase;
- }
-
- /*
* Finally sort ranges by start position. This avoids a dependency on
* keeping features in order on the sequence (if they are in order anyway,
* the sort will have almost no work to do). The implicit assumption is CDS
}
/**
- * 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
- */
- public static int computeProteinFeatures(SequenceI dnaSeq,
- SequenceI peptide, MapList dnaToProtein)
- {
- while (dnaSeq.getDatasetSequence() != null)
- {
- dnaSeq = dnaSeq.getDatasetSequence();
- }
- while (peptide.getDatasetSequence() != null)
- {
- peptide = peptide.getDatasetSequence();
- }
-
- transferFeatures(dnaSeq, peptide, dnaToProtein, SequenceOntologyI.EXON);
-
- /*
- * compute protein variants from dna variants and codon mappings;
- * NB - alternatively we could retrieve this using the REST service e.g.
- * http://rest.ensembl.org/overlap/translation
- * /ENSP00000288602?feature=transcript_variation;content-type=text/xml
- * which would be a bit slower but possibly more reliable
- */
-
- /*
- * build a map with codon variations for each potentially varying peptide
- */
- LinkedHashMap<Integer, List<DnaVariant>[]> variants = buildDnaVariantsMap(
- dnaSeq, dnaToProtein);
-
- /*
- * scan codon variations, compute peptide variants and add to peptide sequence
- */
- int count = 0;
- for (Entry<Integer, List<DnaVariant>[]> variant : variants.entrySet())
- {
- int peptidePos = variant.getKey();
- List<DnaVariant>[] codonVariants = variant.getValue();
- count += computePeptideVariants(peptide, peptidePos, codonVariants);
- }
-
- return count;
- }
-
- /**
- * Computes non-synonymous peptide variants from codon variants and adds them
- * as sequence_variant features on the protein sequence (one feature per
- * allele variant). Selected attributes (variant id, clinical significance)
- * are copied over to the new features.
- *
- * @param peptide
- * the protein sequence
- * @param peptidePos
- * the position to compute peptide variants for
- * @param codonVariants
- * a list of dna variants per codon position
- * @return the number of features added
- */
- static int computePeptideVariants(SequenceI peptide, int peptidePos,
- List<DnaVariant>[] codonVariants)
- {
- String residue = String.valueOf(peptide.getCharAt(peptidePos - 1));
- int count = 0;
- String base1 = codonVariants[0].get(0).base;
- String base2 = codonVariants[1].get(0).base;
- String base3 = codonVariants[2].get(0).base;
-
- /*
- * variants in first codon base
- */
- for (DnaVariant var : codonVariants[0])
- {
- if (var.variant != null)
- {
- String alleles = (String) var.variant.getValue(Gff3Helper.ALLELES);
- if (alleles != null)
- {
- for (String base : alleles.split(","))
- {
- String codon = base + base2 + base3;
- if (addPeptideVariant(peptide, peptidePos, residue, var, codon))
- {
- count++;
- }
- }
- }
- }
- }
-
- /*
- * variants in second codon base
- */
- for (DnaVariant var : codonVariants[1])
- {
- if (var.variant != null)
- {
- String alleles = (String) var.variant.getValue(Gff3Helper.ALLELES);
- if (alleles != null)
- {
- for (String base : alleles.split(","))
- {
- String codon = base1 + base + base3;
- if (addPeptideVariant(peptide, peptidePos, residue, var, codon))
- {
- count++;
- }
- }
- }
- }
- }
-
- /*
- * variants in third codon base
- */
- for (DnaVariant var : codonVariants[2])
- {
- if (var.variant != null)
- {
- String alleles = (String) var.variant.getValue(Gff3Helper.ALLELES);
- if (alleles != null)
- {
- for (String base : alleles.split(","))
- {
- String codon = base1 + base2 + base;
- if (addPeptideVariant(peptide, peptidePos, residue, var, codon))
- {
- count++;
- }
- }
- }
- }
- }
-
- return count;
- }
-
- /**
- * Helper method that adds a peptide variant feature, provided the given codon
- * translates to a value different to the current residue (is a non-synonymous
- * variant). ID and clinical_significance attributes of the dna variant (if
- * present) are copied to the new feature.
- *
- * @param peptide
- * @param peptidePos
- * @param residue
- * @param var
- * @param codon
- * @return true if a feature was added, else false
- */
- static boolean addPeptideVariant(SequenceI peptide, int peptidePos,
- String residue, DnaVariant var, String codon)
- {
- /*
- * get peptide translation of codon e.g. GAT -> D
- * note that variants which are not single alleles,
- * e.g. multibase variants or HGMD_MUTATION etc
- * are currently ignored here
- */
- String trans = codon.contains("-") ? "-"
- : (codon.length() > CODON_LENGTH ? null
- : ResidueProperties.codonTranslate(codon));
- if (trans != null && !trans.equals(residue))
- {
- String residue3Char = StringUtils
- .toSentenceCase(ResidueProperties.aa2Triplet.get(residue));
- String trans3Char = StringUtils
- .toSentenceCase(ResidueProperties.aa2Triplet.get(trans));
- String desc = "p." + residue3Char + peptidePos + trans3Char;
- SequenceFeature sf = new SequenceFeature(
- SequenceOntologyI.SEQUENCE_VARIANT, desc, peptidePos,
- peptidePos, var.getSource());
- StringBuilder attributes = new StringBuilder(32);
- String id = (String) var.variant.getValue(ID);
- if (id != null)
- {
- if (id.startsWith(SEQUENCE_VARIANT))
- {
- id = id.substring(SEQUENCE_VARIANT.length());
- }
- sf.setValue(ID, id);
- attributes.append(ID).append("=").append(id);
- // TODO handle other species variants JAL-2064
- StringBuilder link = new StringBuilder(32);
- try
- {
- link.append(desc).append(" ").append(id).append(
- "|http://www.ensembl.org/Homo_sapiens/Variation/Summary?v=")
- .append(URLEncoder.encode(id, "UTF-8"));
- sf.addLink(link.toString());
- } catch (UnsupportedEncodingException e)
- {
- // as if
- }
- }
- String clinSig = (String) var.variant.getValue(CLINICAL_SIGNIFICANCE);
- if (clinSig != null)
- {
- sf.setValue(CLINICAL_SIGNIFICANCE, clinSig);
- attributes.append(";").append(CLINICAL_SIGNIFICANCE).append("=")
- .append(clinSig);
- }
- peptide.addSequenceFeature(sf);
- if (attributes.length() > 0)
- {
- sf.setAttributes(attributes.toString());
- }
- return true;
- }
- return false;
- }
-
- /**
- * Builds a map whose key is position in the protein sequence, and value is a
- * list of the base and all variants for each corresponding codon position.
- * <p>
- * This depends on dna variants being held as a comma-separated list as
- * property "alleles" on variant features.
- *
- * @param dnaSeq
- * @param dnaToProtein
- * @return
- */
- @SuppressWarnings("unchecked")
- static LinkedHashMap<Integer, List<DnaVariant>[]> buildDnaVariantsMap(
- SequenceI dnaSeq, MapList dnaToProtein)
- {
- /*
- * map from peptide position to all variants of the codon which codes for it
- * LinkedHashMap ensures we keep the peptide features in sequence order
- */
- LinkedHashMap<Integer, List<DnaVariant>[]> variants = new LinkedHashMap<Integer, List<DnaVariant>[]>();
-
- List<SequenceFeature> dnaFeatures = dnaSeq.getFeatures()
- .getFeaturesByOntology(SequenceOntologyI.SEQUENCE_VARIANT);
- if (dnaFeatures.isEmpty())
- {
- 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;
- }
-
- /*
- * ignore variant if not a SNP
- */
- String alls = (String) sf.getValue(Gff3Helper.ALLELES);
- if (alls == null)
- {
- continue; // non-SNP VCF variant perhaps - can't process this
- }
-
- String[] alleles = alls.toUpperCase().split(",");
- boolean isSnp = true;
- for (String allele : alleles)
- {
- if (allele.trim().length() > 1)
- {
- isSnp = false;
- }
- }
- if (!isSnp)
- {
- continue;
- }
-
- int[] mapsTo = dnaToProtein.locateInTo(dnaCol, dnaCol);
- if (mapsTo == null)
- {
- // feature doesn't lie within coding region
- continue;
- }
- int peptidePosition = mapsTo[0];
- List<DnaVariant>[] codonVariants = variants.get(peptidePosition);
- if (codonVariants == null)
- {
- codonVariants = new ArrayList[CODON_LENGTH];
- codonVariants[0] = new ArrayList<DnaVariant>();
- codonVariants[1] = new ArrayList<DnaVariant>();
- codonVariants[2] = new ArrayList<DnaVariant>();
- variants.put(peptidePosition, codonVariants);
- }
-
- /*
- * get this peptide's 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 any variant) for each codon position
- */
- for (int codonPos = 0; codonPos < CODON_LENGTH; codonPos++)
- {
- String nucleotide = String.valueOf(
- dnaSeq.getCharAt(codon[codonPos] - dnaStart)).toUpperCase();
- List<DnaVariant> codonVariant = codonVariants[codonPos];
- if (codon[codonPos] == dnaCol)
- {
- if (!codonVariant.isEmpty()
- && codonVariant.get(0).variant == null)
- {
- /*
- * already recorded base value, add this variant
- */
- codonVariant.get(0).variant = sf;
- }
- else
- {
- /*
- * add variant with base value
- */
- codonVariant.add(new DnaVariant(nucleotide, sf));
- }
- }
- else if (codonVariant.isEmpty())
- {
- /*
- * record (possibly non-varying) base value
- */
- codonVariant.add(new DnaVariant(nucleotide));
- }
- }
- }
- return variants;
- }
-
- /**
* Makes an alignment with a copy of the given sequences, adding in any
* non-redundant sequences which are mapped to by the cross-referenced
* sequences.
/*
* fancy case - aligning via mappings between sequences
*/
- List<SequenceI> unmapped = new ArrayList<SequenceI>();
+ List<SequenceI> unmapped = new ArrayList<>();
Map<Integer, Map<SequenceI, Character>> columnMap = buildMappedColumnsMap(
unaligned, aligned, unmapped);
int width = columnMap.size();
* true; else returns false
*
* @param unaligned
- * - sequences to be aligned based on aligned
+ * - sequences to be aligned based on aligned
* @param aligned
- * - 'guide' alignment containing sequences derived from same dataset
- * as unaligned
+ * - 'guide' alignment containing sequences derived from same
+ * dataset as unaligned
* @return
*/
static boolean alignAsSameSequences(AlignmentI unaligned,
}
// map from dataset sequence to alignment sequence(s)
- Map<SequenceI, List<SequenceI>> alignedDatasets = new HashMap<SequenceI, List<SequenceI>>();
+ Map<SequenceI, List<SequenceI>> alignedDatasets = new HashMap<>();
for (SequenceI seq : aligned.getSequences())
{
SequenceI ds = seq.getDatasetSequence();
}
/*
- * first pass - check whether all sequences to be aligned share a dataset
- * sequence with an aligned sequence
+ * first pass - check whether all sequences to be aligned share a
+ * dataset sequence with an aligned sequence; also note the leftmost
+ * ungapped column from which to copy
*/
+ int leftmost = Integer.MAX_VALUE;
for (SequenceI seq : unaligned.getSequences())
{
- if (!alignedDatasets.containsKey(seq.getDatasetSequence()))
+ final SequenceI ds = seq.getDatasetSequence();
+ if (!alignedDatasets.containsKey(ds))
{
return false;
}
+ SequenceI alignedSeq = alignedDatasets.get(ds)
+ .get(0);
+ int startCol = alignedSeq.findIndex(seq.getStart()); // 1..
+ leftmost = Math.min(leftmost, startCol);
}
/*
* heuristic rule: pair off sequences in order for the case where
* more than one shares the same dataset sequence
*/
+ final char gapCharacter = aligned.getGapCharacter();
for (SequenceI seq : unaligned.getSequences())
{
List<SequenceI> alignedSequences = alignedDatasets
.get(seq.getDatasetSequence());
- // TODO: getSequenceAsString() will be deprecated in the future
- // TODO: need to leave to SequenceI implementor to update gaps
- seq.setSequence(alignedSequences.get(0).getSequenceAsString());
+ if (alignedSequences.isEmpty())
+ {
+ /*
+ * defensive check - shouldn't happen! (JAL-3536)
+ */
+ continue;
+ }
+ SequenceI alignedSeq = alignedSequences.get(0);
+
+ /*
+ * gap fill for leading (5') UTR if any
+ */
+ // TODO this copies intron columns - wrong!
+ int startCol = alignedSeq.findIndex(seq.getStart()); // 1..
+ int endCol = alignedSeq.findIndex(seq.getEnd());
+ char[] seqchars = new char[endCol - leftmost + 1];
+ Arrays.fill(seqchars, gapCharacter);
+ char[] toCopy = alignedSeq.getSequence(startCol - 1, endCol);
+ System.arraycopy(toCopy, 0, seqchars, startCol - leftmost,
+ toCopy.length);
+ seq.setSequence(String.valueOf(seqchars));
if (alignedSequences.size() > 0)
{
// pop off aligned sequences (except the last one)
}
}
+ /*
+ * finally remove gapped columns (e.g. introns)
+ */
+ new RemoveGapColCommand("", unaligned.getSequencesArray(), 0,
+ unaligned.getWidth() - 1, unaligned);
+
return true;
}
* {unalignedSequence, characterPerSequence} at that position.
* TreeMap keeps the entries in ascending column order.
*/
- SortedMap<Integer, Map<SequenceI, Character>> map = new TreeMap<Integer, Map<SequenceI, Character>>();
+ SortedMap<Integer, Map<SequenceI, Character>> map = new TreeMap<>();
/*
* record any sequences that have no mapping so can't be realigned
Map<SequenceI, Character> seqsMap = map.get(fromCol);
if (seqsMap == null)
{
- seqsMap = new HashMap<SequenceI, Character>();
+ seqsMap = new HashMap<>();
map.put(fromCol, seqsMap);
}
seqsMap.put(seq, seq.getCharAt(mappedCharPos - toStart));
git://source.jalview.org / jalview.git / blobdiff