- * 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);
- }
-
- /*
- * sort to get sequence features in start position order
- * - would be better to store in Sequence as a TreeSet or NCList?
- */
- if (peptide.getSequenceFeatures() != null)
- {
- Arrays.sort(peptide.getSequenceFeatures(),
- new Comparator<SequenceFeature>()
- {
- @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;
- }
-
- /**
- * 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("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("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("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.