From 246a3d3d91e543122a5dcbebfa0b5d29d7854aa5 Mon Sep 17 00:00:00 2001 From: gmungoc Date: Thu, 28 Jan 2021 16:56:55 +0000 Subject: [PATCH] JAL-3806 (2.11.2 version) more lax allowance (3 positions) for stop codon --- src/jalview/datamodel/AlignedCodonFrame.java | 137 ++++++++++++++++++++++++++ 1 file changed, 137 insertions(+) diff --git a/src/jalview/datamodel/AlignedCodonFrame.java b/src/jalview/datamodel/AlignedCodonFrame.java index fffa137..3c33063 100644 --- a/src/jalview/datamodel/AlignedCodonFrame.java +++ b/src/jalview/datamodel/AlignedCodonFrame.java @@ -107,6 +107,143 @@ public class AlignedCodonFrame { return mapping; } + + /** + * Returns true if the mapping covers the full length of the given sequence. + * This allows us to distinguish the CDS that codes for a protein from + * another overlapping CDS in the parent dna sequence. + * + * @param seq + * @return + */ + public boolean covers(SequenceI seq) + { + return covers(seq,false,false); + } + /** + * + * @param seq + * @param localCover - when true - compare extent of seq's dataset sequence rather than the local extent + * @param either - when true coverage is required for either seq or the mapped sequence + * @return true if mapping covers full length of given sequence (or the other if either==true) + */ + public boolean covers(SequenceI seq, boolean localCover,boolean either) + { + List mappedRanges = null,otherRanges=null; + MapList mapList = mapping.getMap(); + int mstart=seq.getStart(),mend=seq.getEnd(),ostart,oend; + ; + if (fromSeq == seq || fromSeq == seq.getDatasetSequence()) + { + if (localCover && fromSeq !=seq) + { + mstart=fromSeq.getStart(); + mend=fromSeq.getEnd(); + } + mappedRanges = mapList.getFromRanges(); + otherRanges=mapList.getToRanges(); + ostart=mapping.to.getStart(); + oend=mapping.to.getEnd(); + } + else if (mapping.to == seq || mapping.to == seq.getDatasetSequence()) + { + if (localCover && mapping.to !=seq) + { + mstart=mapping.to.getStart(); + mend=mapping.to.getEnd(); + } + mappedRanges = mapList.getToRanges(); + otherRanges=mapList.getFromRanges(); + ostart=fromSeq.getStart(); + oend=fromSeq.getEnd(); + } + else + { + return false; + } + + /* + * check that each mapped range lies within the sequence range + * (necessary for circular CDS - example EMBL:J03321:AAA91567) + * and mapped length covers (at least) sequence length + */ + int length = countRange(mappedRanges,mstart,mend); + + if (length != -1) + { + // add 3 to mapped length to allow for a mapped stop codon + if (length + 3 >= (mend - mstart + 1)) + { + return true; + } + } + if (either) + { + // also check coverage of the other range + length = countRange(otherRanges, ostart, oend); + if (length != -1) + { + if (length + 1 >= (oend - ostart + 1)) + { + return true; + } + } + } + return false; + } + private int countRange(List mappedRanges,int mstart,int mend) + { + int length=0; + for (int[] range : mappedRanges) + { + int from = Math.min(range[0], range[1]); + int to = Math.max(range[0], range[1]); + if (from < mstart || to > mend) + { + return -1; + } + length += (to - from + 1); + } + return length; + } + + /** + * Adds any regions mapped to or from position {@code pos} in sequence + * {@code seq} to the given search results + * + * @param seq + * @param pos + * @param sr + */ + public void markMappedRegion(SequenceI seq, int pos, SearchResultsI sr) + { + int[] codon = null; + SequenceI mappedSeq = null; + SequenceI ds = seq.getDatasetSequence(); + if (ds == null) + { + ds = seq; + } + + if (this.fromSeq == seq || this.fromSeq == ds) + { + codon = this.mapping.map.locateInTo(pos, pos); + mappedSeq = this.mapping.to; + } + else if (this.mapping.to == seq || this.mapping.to == ds) + { + codon = this.mapping.map.locateInFrom(pos, pos); + mappedSeq = this.fromSeq; + } + + if (codon != null) + { + for (int i = 0; i < codon.length; i += 2) + { + sr.addResult(mappedSeq, codon[i], codon[i + 1]); + } + } + } } private List mappings; -- 1.7.10.2