X-Git-Url: http://source.jalview.org/gitweb/?a=blobdiff_plain;f=src%2Fjalview%2Fdatamodel%2FAlignedCodonFrame.java;h=0927dda472b244a41fd8c2d5c293f5fb64503aa0;hb=87a707afdbbe1675330f7e82faaf9443af89cefe;hp=9fa11ccd98c9cd4212a04970920ac23a574347a5;hpb=c9296c060f69f3eeab1f6da2a07dbdb87c946e52;p=jalview.git diff --git a/src/jalview/datamodel/AlignedCodonFrame.java b/src/jalview/datamodel/AlignedCodonFrame.java index 9fa11cc..0927dda 100644 --- a/src/jalview/datamodel/AlignedCodonFrame.java +++ b/src/jalview/datamodel/AlignedCodonFrame.java @@ -20,13 +20,13 @@ */ package jalview.datamodel; -import jalview.util.MapList; -import jalview.util.MappingUtils; - import java.util.AbstractList; import java.util.ArrayList; import java.util.List; +import jalview.util.MapList; +import jalview.util.MappingUtils; + /** * Stores mapping between the columns of a protein alignment and a DNA alignment * and a list of individual codon to amino acid mappings between sequences. @@ -118,19 +118,20 @@ public class AlignedCodonFrame */ public boolean covers(SequenceI seq) { - return covers(seq,false); + return covers(seq,false,false); } /** * * @param seq - * @param localCover - when true - compare extent of seq's dataset sequence rather than the local extent - * @return true if mapping covers full length of given sequence + * @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) + public boolean covers(SequenceI seq, boolean localCover,boolean either) { - List mappedRanges = null; + List mappedRanges = null,otherRanges=null; MapList mapList = mapping.getMap(); - int mstart=seq.getStart(),mend=seq.getEnd() + int mstart=seq.getStart(),mend=seq.getEnd(),ostart,oend; ; if (fromSeq == seq || fromSeq == seq.getDatasetSequence()) { @@ -140,6 +141,9 @@ public class AlignedCodonFrame 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()) { @@ -149,6 +153,9 @@ public class AlignedCodonFrame mend=mapping.to.getEnd(); } mappedRanges = mapList.getToRanges(); + otherRanges=mapList.getFromRanges(); + ostart=fromSeq.getStart(); + oend=fromSeq.getEnd(); } else { @@ -156,27 +163,82 @@ public class AlignedCodonFrame } /* - * check that each mapped range lieS with the sequence range + * 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 = 0; + int length = countRange(mappedRanges,mstart,mend); + + if (length != -1) + { + // add 1 to mapped length to allow for a mapped stop codon + if (length + 1 >= (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 false; + return -1; } length += (to - from + 1); } - // add 1 to mapped length to allow for a mapped stop codon - if (length + 1 < (mend - mstart + 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 (this.fromSeq == seq || this.fromSeq == ds) { - return false; + 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]); + } } - return true; } } @@ -378,34 +440,10 @@ public class AlignedCodonFrame public void markMappedRegion(SequenceI seq, int index, SearchResultsI results) { - int[] codon; SequenceI ds = seq.getDatasetSequence(); for (SequenceToSequenceMapping ssm : mappings) { - if (ssm.fromSeq == seq || ssm.fromSeq == ds) - { - codon = ssm.mapping.map.locateInTo(index, index); - if (codon != null) - { - for (int i = 0; i < codon.length; i += 2) - { - results.addResult(ssm.mapping.to, codon[i], codon[i + 1]); - } - } - } - else if (ssm.mapping.to == seq || ssm.mapping.to == ds) - { - { - codon = ssm.mapping.map.locateInFrom(index, index); - if (codon != null) - { - for (int i = 0; i < codon.length; i += 2) - { - results.addResult(ssm.fromSeq, codon[i], codon[i + 1]); - } - } - } - } + ssm.markMappedRegion(ds, index, results); } } @@ -455,18 +493,39 @@ public class AlignedCodonFrame */ public SequenceI findAlignedSequence(SequenceI seq, AlignmentI al) { + return findAlignedSequence(seq, al, null); + } + /** + * Convenience method to return the first aligned sequence in the given + * alignment whose dataset has a mapping with the given (aligned or dataset) + * sequence, and optionally the mapping that relates them + * + * @param seq + * @param al + * @param map - list to add the mapping to + * @return sequence from al that maps to seq + */ + public SequenceI findAlignedSequence(SequenceI seq, AlignmentI al,List map) + { /* * Search mapped protein ('to') sequences first. */ for (SequenceToSequenceMapping ssm : mappings) { - if (ssm.fromSeq == seq || ssm.fromSeq == seq.getDatasetSequence()) + int mStart=ssm.getMapping().getMap().getFromLowest(),mEnd=ssm.getMapping().map.getFromHighest(); + if ((ssm.fromSeq == seq || ssm.fromSeq == seq.getDatasetSequence()) + // here AlignmentUtilsTest. testAlignProteinAsDna_incompleteStartCodon fails because mStart/mEnd is contained by seq + // without this filter, we don't get the correct mapping, however + )// && seq.getStart()>=mStart && seq.getEnd()<=mEnd) { for (SequenceI sourceAligned : al.getSequences()) { - if (ssm.mapping.to == sourceAligned.getDatasetSequence() - || ssm.mapping.to == sourceAligned) + if (ssm.covers(sourceAligned,true,false)) { + if (map != null) + { + map.add(ssm); + } return sourceAligned; } } @@ -478,13 +537,19 @@ public class AlignedCodonFrame */ for (SequenceToSequenceMapping ssm : mappings) { - if (ssm.mapping.to == seq + int mStart=ssm.getMapping().getMap().getToLowest(),mEnd=ssm.getMapping().map.getToHighest(); + if ((ssm.mapping.to == seq || ssm.mapping.to == seq.getDatasetSequence()) + && seq.getStart()>=mStart && seq.getEnd()<=mEnd) { for (SequenceI sourceAligned : al.getSequences()) { - if (ssm.fromSeq == sourceAligned.getDatasetSequence()) + if (ssm.covers(sourceAligned,true,true)) { + if (map != null) + { + map.add(ssm); + } return sourceAligned; } } @@ -878,4 +943,34 @@ public class AlignedCodonFrame } return null; } + + /** + * Returns the first mapping found which is between the given sequence and + * another, is a triplet mapping (3:1 or 1:3), and covers the full extent of + * both sequences involved. + * + * @param seq + * @return + */ + public SequenceToSequenceMapping getCoveringCodonMapping(SequenceI seq) + { + for (SequenceToSequenceMapping mapping : mappings) + { + if (mapping.getMapping().getMap().isTripletMap() + && mapping.covers(seq)) + { + if (mapping.fromSeq == seq + && mapping.covers(mapping.getMapping().getTo())) + { + return mapping; + } + else if (mapping.getMapping().getTo() == seq + && mapping.covers(mapping.fromSeq)) + { + return mapping; + } + } + } + return null; + } }