- * 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,
- SequenceOntology.EXON);
-
- LinkedHashMap<Integer, String[][]> variants = buildDnaVariantsMap(
- dnaSeq, dnaToProtein);
-
- /*
- * scan codon variations, compute peptide variants and add to peptide sequence
- */
- int count = 0;
- for (Entry<Integer, String[][]> variant : variants.entrySet())
- {
- int peptidePos = variant.getKey();
- String[][] codonVariants = variant.getValue();
- String residue = String.valueOf(peptide.getCharAt(peptidePos - 1)); // 0-based
- List<String> peptideVariants = computePeptideVariants(codonVariants,
- residue);
- if (!peptideVariants.isEmpty())
- {
- String desc = StringUtils.listToDelimitedString(peptideVariants,
- ", ");
- SequenceFeature sf = new SequenceFeature(
- SequenceOntology.SEQUENCE_VARIANT, desc, peptidePos,
- peptidePos, 0f, null);
- peptide.addSequenceFeature(sf);
- count++;
- }
- }
- 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<Integer, String[][]> 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<Integer, String[][]> variants = new LinkedHashMap<Integer, String[][]>();
- SequenceOntology so = SequenceOntology.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.isSequenceVariant(sf.getType()))
- {
- 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<String> computePeptideVariants(
- String[][] codonVariants, String residue)
- {
- List<String> result = new ArrayList<String>();
- for (String base1 : codonVariants[0])
- {
- for (String base2 : codonVariants[1])
- {
- 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);
- }
- }
- }
- }
-
- /*
- * sort alphabetically with STOP at the end
- */
- Collections.sort(result, new Comparator<String>()
- {
-
- @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;
- }
-
- /**