2 * Jalview - A Sequence Alignment Editor and Viewer (Version 2.8.2)
3 * Copyright (C) 2014 The Jalview Authors
5 * This file is part of Jalview.
7 * Jalview is free software: you can redistribute it and/or
8 * modify it under the terms of the GNU General Public License
9 * as published by the Free Software Foundation, either version 3
10 * of the License, or (at your option) any later version.
12 * Jalview is distributed in the hope that it will be useful, but
13 * WITHOUT ANY WARRANTY; without even the implied warranty
14 * of MERCHANTABILITY or FITNESS FOR A PARTICULAR
15 * PURPOSE. See the GNU General Public License for more details.
17 * You should have received a copy of the GNU General Public License
18 * along with Jalview. If not, see <http://www.gnu.org/licenses/>.
19 * The Jalview Authors are detailed in the 'AUTHORS' file.
21 package jalview.analysis;
23 import java.util.ArrayList;
24 import java.util.Hashtable;
25 import java.util.Vector;
27 import jalview.datamodel.AlignedCodonFrame;
28 import jalview.datamodel.Alignment;
29 import jalview.datamodel.AlignmentAnnotation;
30 import jalview.datamodel.AlignmentI;
31 import jalview.datamodel.Annotation;
32 import jalview.datamodel.DBRefEntry;
33 import jalview.datamodel.FeatureProperties;
34 import jalview.datamodel.Mapping;
35 import jalview.datamodel.Sequence;
36 import jalview.datamodel.SequenceFeature;
37 import jalview.datamodel.SequenceI;
38 import jalview.schemes.ResidueProperties;
39 import jalview.util.MapList;
40 import jalview.util.ShiftList;
48 * @return -1 if cdp1 aligns before cdp2, 0 if in the same column or cdp2 is
49 * null, +1 if after cdp2
51 private static int compare_codonpos(int[] cdp1, int[] cdp2)
54 || (cdp1[0] == cdp2[0] && cdp1[1] == cdp2[1] && cdp1[2] == cdp2[2]))
56 if (cdp1[0] < cdp2[0] || cdp1[1] < cdp2[1] || cdp1[2] < cdp2[2])
57 return -1; // one base in cdp1 precedes the corresponding base in the
59 return 1; // one base in cdp1 appears after the corresponding base in the
64 * DNA->mapped protein sequence alignment translation given set of sequences
65 * 1. id distinct coding regions within selected region for each sequence 2.
66 * generate peptides based on inframe (or given) translation or (optionally
67 * and where specified) out of frame translations (annotated appropriately) 3.
68 * align peptides based on codon alignment
71 * id potential products from dna 1. search for distinct products within
72 * selected region for each selected sequence 2. group by associated DB type.
73 * 3. return as form for input into above function
79 * create a new alignment of protein sequences by an inframe translation of
80 * the provided NA sequences
89 * destination dataset for translated sequences and mappings
92 public static AlignmentI CdnaTranslate(SequenceI[] selection,
93 String[] seqstring, int viscontigs[], char gapCharacter,
94 AlignmentAnnotation[] annotations, int aWidth, Alignment dataset)
96 return CdnaTranslate(selection, seqstring, null, viscontigs,
97 gapCharacter, annotations, aWidth, dataset);
105 * - array of DbRefEntry objects from which exon map in seqstring is
108 * @param gapCharacter
114 public static AlignmentI CdnaTranslate(SequenceI[] selection,
115 String[] seqstring, DBRefEntry[] product, int viscontigs[],
116 char gapCharacter, AlignmentAnnotation[] annotations, int aWidth,
119 AlignedCodonFrame codons = new AlignedCodonFrame(aWidth); // stores hash of
125 int s, sSize = selection.length;
126 Vector pepseqs = new Vector();
127 for (s = 0; s < sSize; s++)
129 SequenceI newseq = translateCodingRegion(selection[s], seqstring[s],
130 viscontigs, codons, gapCharacter,
131 (product != null) ? product[s] : null, false); // possibly
136 pepseqs.addElement(newseq);
137 SequenceI ds = newseq;
140 while (ds.getDatasetSequence() != null)
142 ds = ds.getDatasetSequence();
144 dataset.addSequence(ds);
148 if (codons.aaWidth == 0)
150 SequenceI[] newseqs = new SequenceI[pepseqs.size()];
151 pepseqs.copyInto(newseqs);
152 AlignmentI al = new Alignment(newseqs);
153 al.padGaps(); // ensure we look aligned.
154 al.setDataset(dataset);
155 translateAlignedAnnotations(annotations, al, codons);
156 al.addCodonFrame(codons);
161 * fake the collection of DbRefs with associated exon mappings to identify if
162 * a translation would generate distinct product in the currently selected
169 public static boolean canTranslate(SequenceI[] selection,
172 for (int gd = 0; gd < selection.length; gd++)
174 SequenceI dna = selection[gd];
175 jalview.datamodel.DBRefEntry[] dnarefs = jalview.util.DBRefUtils
176 .selectRefs(dna.getDBRef(),
177 jalview.datamodel.DBRefSource.DNACODINGDBS);
180 // intersect with pep
181 // intersect with pep
182 Vector mappedrefs = new Vector();
183 DBRefEntry[] refs = dna.getDBRef();
184 for (int d = 0; d < refs.length; d++)
186 if (refs[d].getMap() != null && refs[d].getMap().getMap() != null
187 && refs[d].getMap().getMap().getFromRatio() == 3
188 && refs[d].getMap().getMap().getToRatio() == 1)
190 mappedrefs.addElement(refs[d]); // add translated protein maps
193 dnarefs = new DBRefEntry[mappedrefs.size()];
194 mappedrefs.copyInto(dnarefs);
195 for (int d = 0; d < dnarefs.length; d++)
197 Mapping mp = dnarefs[d].getMap();
200 for (int vc = 0; vc < viscontigs.length; vc += 2)
202 int[] mpr = mp.locateMappedRange(viscontigs[vc],
217 * generate a set of translated protein products from annotated sequenceI
221 * @param gapCharacter
223 * destination dataset for translated sequences
228 public static AlignmentI CdnaTranslate(SequenceI[] selection,
229 int viscontigs[], char gapCharacter, Alignment dataset)
232 Vector cdnasqs = new Vector();
233 Vector cdnasqi = new Vector();
234 Vector cdnaprod = new Vector();
235 for (int gd = 0; gd < selection.length; gd++)
237 SequenceI dna = selection[gd];
238 jalview.datamodel.DBRefEntry[] dnarefs = jalview.util.DBRefUtils
239 .selectRefs(dna.getDBRef(),
240 jalview.datamodel.DBRefSource.DNACODINGDBS);
243 // intersect with pep
244 Vector mappedrefs = new Vector();
245 DBRefEntry[] refs = dna.getDBRef();
246 for (int d = 0; d < refs.length; d++)
248 if (refs[d].getMap() != null && refs[d].getMap().getMap() != null
249 && refs[d].getMap().getMap().getFromRatio() == 3
250 && refs[d].getMap().getMap().getToRatio() == 1)
252 mappedrefs.addElement(refs[d]); // add translated protein maps
255 dnarefs = new DBRefEntry[mappedrefs.size()];
256 mappedrefs.copyInto(dnarefs);
257 for (int d = 0; d < dnarefs.length; d++)
259 Mapping mp = dnarefs[d].getMap();
260 StringBuffer sqstr = new StringBuffer();
263 Mapping intersect = mp.intersectVisContigs(viscontigs);
264 // generate seqstring for this sequence based on mapping
266 if (sqstr.length() > alwidth)
267 alwidth = sqstr.length();
268 cdnasqs.addElement(sqstr.toString());
269 cdnasqi.addElement(dna);
270 cdnaprod.addElement(intersect);
274 SequenceI[] cdna = new SequenceI[cdnasqs.size()];
275 DBRefEntry[] prods = new DBRefEntry[cdnaprod.size()];
276 String[] xons = new String[cdnasqs.size()];
277 cdnasqs.copyInto(xons);
278 cdnaprod.copyInto(prods);
279 cdnasqi.copyInto(cdna);
280 return CdnaTranslate(cdna, xons, prods, viscontigs, gapCharacter,
281 null, alwidth, dataset);
287 * translate na alignment annotations onto translated amino acid alignment al
288 * using codon mapping codons
294 public static void translateAlignedAnnotations(
295 AlignmentAnnotation[] annotations, AlignmentI al,
296 AlignedCodonFrame codons)
298 // //////////////////////////////
299 // Copy annotations across
301 // Can only do this for columns with consecutive codons, or where
302 // annotation is sequence associated.
305 if (annotations != null)
307 for (int i = 0; i < annotations.length; i++)
309 // Skip any autogenerated annotation
310 if (annotations[i].autoCalculated)
315 aSize = codons.getaaWidth(); // aa alignment width.
316 jalview.datamodel.Annotation[] anots = (annotations[i].annotations == null) ? null
317 : new jalview.datamodel.Annotation[aSize];
320 for (a = 0; a < aSize; a++)
322 // process through codon map.
323 if (codons.codons[a] != null
324 && codons.codons[a][0] == (codons.codons[a][2] - 2))
326 anots[a] = getCodonAnnotation(codons.codons[a],
327 annotations[i].annotations);
332 jalview.datamodel.AlignmentAnnotation aa = new jalview.datamodel.AlignmentAnnotation(
333 annotations[i].label, annotations[i].description, anots);
334 aa.graph = annotations[i].graph;
335 aa.graphGroup = annotations[i].graphGroup;
336 aa.graphHeight = annotations[i].graphHeight;
337 if (annotations[i].getThreshold() != null)
339 aa.setThreshold(new jalview.datamodel.GraphLine(annotations[i]
342 if (annotations[i].hasScore)
344 aa.setScore(annotations[i].getScore());
346 if (annotations[i].sequenceRef != null)
348 SequenceI aaSeq = codons
349 .getAaForDnaSeq(annotations[i].sequenceRef);
352 // aa.compactAnnotationArray(); // throw away alignment annotation
354 aa.setSequenceRef(aaSeq);
355 aa.createSequenceMapping(aaSeq, aaSeq.getStart(), true); // rebuild
357 aa.adjustForAlignment();
358 aaSeq.addAlignmentAnnotation(aa);
362 al.addAnnotation(aa);
367 private static Annotation getCodonAnnotation(int[] is,
368 Annotation[] annotations)
370 // Have a look at all the codon positions for annotation and put the first
371 // one found into the translated annotation pos.
373 Annotation annot = null;
374 for (int p = 0; p < 3; p++)
376 if (annotations[is[p]] != null)
380 annot = new Annotation(annotations[is[p]]);
386 Annotation cpy = new Annotation(annotations[is[p]]);
387 if (annot.colour == null)
389 annot.colour = cpy.colour;
391 if (annot.description == null || annot.description.length() == 0)
393 annot.description = cpy.description;
395 if (annot.displayCharacter == null)
397 annot.displayCharacter = cpy.displayCharacter;
399 if (annot.secondaryStructure == 0)
401 annot.secondaryStructure = cpy.secondaryStructure;
403 annot.value += cpy.value;
410 annot.value /= (float) contrib;
416 * Translate a na sequence
419 * sequence displayed under viscontigs visible columns
421 * ORF read in some global alignment reference frame
423 * mapping from global reference frame to visible seqstring ORF read
425 * Definition of global ORF alignment reference frame
426 * @param gapCharacter
427 * @return sequence ready to be added to alignment.
429 * {@link #translateCodingRegion(SequenceI,String,int[],AlignedCodonFrame,char,DBRefEntry,boolean)}
432 public static SequenceI translateCodingRegion(SequenceI selection,
433 String seqstring, int[] viscontigs, AlignedCodonFrame codons,
434 char gapCharacter, DBRefEntry product)
436 return translateCodingRegion(selection, seqstring, viscontigs, codons,
437 gapCharacter, product, false);
441 * Translate a na sequence
444 * sequence displayed under viscontigs visible columns
446 * ORF read in some global alignment reference frame
448 * mapping from global reference frame to visible seqstring ORF read
450 * Definition of global ORF alignment reference frame
451 * @param gapCharacter
453 * when true stop codons will translate as '*', otherwise as 'X'
454 * @return sequence ready to be added to alignment.
456 public static SequenceI translateCodingRegion(SequenceI selection,
457 String seqstring, int[] viscontigs, AlignedCodonFrame codons,
458 char gapCharacter, DBRefEntry product, final boolean starForStop)
460 java.util.List skip = new ArrayList();
461 int skipint[] = null;
462 ShiftList vismapping = new ShiftList(); // map from viscontigs to seqstring
464 int vc, scontigs[] = new int[viscontigs.length];
466 for (vc = 0; vc < viscontigs.length; vc += 2)
470 vismapping.addShift(npos, viscontigs[vc]);
475 vismapping.addShift(npos, viscontigs[vc] - viscontigs[vc - 1] + 1);
477 scontigs[vc] = viscontigs[vc];
478 scontigs[vc + 1] = viscontigs[vc + 1];
481 StringBuffer protein = new StringBuffer();
482 String seq = seqstring.replace('U', 'T');
483 char codon[] = new char[3];
484 int cdp[] = new int[3], rf = 0, lastnpos = 0, nend;
487 for (npos = 0, nend = seq.length(); npos < nend; npos++)
489 if (!jalview.util.Comparison.isGap(seq.charAt(npos)))
491 cdp[rf] = npos; // store position
492 codon[rf++] = seq.charAt(npos); // store base
494 // filled an RF yet ?
497 String aa = ResidueProperties.codonTranslate(new String(codon));
501 aa = String.valueOf(gapCharacter);
514 skipint[0] = vismapping.shift(skipint[0]);
515 skipint[1] = vismapping.shift(skipint[1]);
516 for (vc = 0; vc < scontigs.length;)
518 if (scontigs[vc + 1] < skipint[0])
520 // before skipint starts
524 if (scontigs[vc] > skipint[1])
526 // finished editing so
529 // Edit the contig list to include the skipped region which did
532 // from : s1 e1 s2 e2 s3 e3
533 // to s: s1 e1 s2 k0 k1 e2 s3 e3
534 // list increases by one unless one boundary (s2==k0 or e2==k1)
535 // matches, and decreases by one if skipint intersects whole
537 if (scontigs[vc] <= skipint[0])
539 if (skipint[0] == scontigs[vc])
541 // skipint at start of contig
542 // shift the start of this contig
543 if (scontigs[vc + 1] > skipint[1])
545 scontigs[vc] = skipint[1];
550 if (scontigs[vc + 1] == skipint[1])
553 t = new int[scontigs.length - 2];
556 System.arraycopy(scontigs, 0, t, 0, vc - 1);
558 if (vc + 2 < t.length)
560 System.arraycopy(scontigs, vc + 2, t, vc, t.length
567 // truncate contig to before the skipint region
568 scontigs[vc + 1] = skipint[0] - 1;
575 // scontig starts before start of skipint
576 if (scontigs[vc + 1] < skipint[1])
578 // skipint truncates end of scontig
579 scontigs[vc + 1] = skipint[0] - 1;
584 // divide region to new contigs
585 t = new int[scontigs.length + 2];
586 System.arraycopy(scontigs, 0, t, 0, vc + 1);
587 t[vc + 1] = skipint[0];
588 t[vc + 2] = skipint[1];
589 System.arraycopy(scontigs, vc + 1, t, vc + 3,
590 scontigs.length - (vc + 1));
600 if (aa.equals("STOP"))
602 aa = starForStop ? "*" : "X";
606 // insert/delete gaps prior to this codon - if necessary
607 boolean findpos = true;
610 // first ensure that the codons array is long enough.
611 codons.checkCodonFrameWidth(aspos);
612 // now check to see if we place the aa at the current aspos in the
614 switch (Dna.compare_codonpos(cdp, codons.codons[aspos]))
617 codons.insertAAGap(aspos, gapCharacter);
621 // this aa appears after the aligned codons at aspos, so prefix it
623 aa = "" + gapCharacter + aa;
625 // if (aspos >= codons.aaWidth)
626 // codons.aaWidth = aspos + 1;
627 break; // check the next position for alignment
629 // codon aligns at aspos position.
633 // codon aligns with all other sequence residues found at aspos
636 if (codons.codons[aspos] == null)
638 // mark this column as aligning to this aligned reading frame
639 codons.codons[aspos] = new int[]
640 { cdp[0], cdp[1], cdp[2] };
642 if (aspos >= codons.aaWidth)
644 // update maximum alignment width
645 // (we can do this without calling checkCodonFrameWidth because it was
646 // already done above)
647 codons.setAaWidth(aspos);
649 // ready for next translated reading frame alignment position (if any)
655 SequenceI newseq = new Sequence(selection.getName(),
659 if (jalview.bin.Cache.log != null)
661 jalview.bin.Cache.log
662 .debug("trimming contigs for incomplete terminal codon.");
667 .println("trimming contigs for incomplete terminal codon.");
669 // map and trim contigs to ORF region
670 vc = scontigs.length - 1;
671 lastnpos = vismapping.shift(lastnpos); // place npos in context of
672 // whole dna alignment (rather
673 // than visible contigs)
674 // incomplete ORF could be broken over one or two visible contig
676 while (vc >= 0 && scontigs[vc] > lastnpos)
678 if (vc > 0 && scontigs[vc - 1] > lastnpos)
684 // correct last interval in list.
685 scontigs[vc] = lastnpos;
689 if (vc > 0 && (vc + 1) < scontigs.length)
691 // truncate map list to just vc elements
692 int t[] = new int[vc + 1];
693 System.arraycopy(scontigs, 0, t, 0, vc + 1);
699 if (scontigs != null)
702 // map scontigs to actual sequence positions on selection
703 for (vc = 0; vc < scontigs.length; vc += 2)
705 scontigs[vc] = selection.findPosition(scontigs[vc]); // not from 1!
706 scontigs[vc + 1] = selection.findPosition(scontigs[vc + 1]); // exclusive
707 if (scontigs[vc + 1] == selection.getEnd())
710 // trim trailing empty intervals.
711 if ((vc + 2) < scontigs.length)
713 int t[] = new int[vc + 2];
714 System.arraycopy(scontigs, 0, t, 0, vc + 2);
718 * delete intervals in scontigs which are not translated. 1. map skip
719 * into sequence position intervals 2. truncate existing ranges and add
720 * new ranges to exclude untranslated regions. if (skip.size()>0) {
721 * Vector narange = new Vector(); for (vc=0; vc<scontigs.length; vc++) {
722 * narange.addElement(new int[] {scontigs[vc]}); } int sint=0,iv[]; vc =
723 * 0; while (sint<skip.size()) { skipint = (int[]) skip.elementAt(sint);
724 * do { iv = (int[]) narange.elementAt(vc); if (iv[0]>=skipint[0] &&
725 * iv[0]<=skipint[1]) { if (iv[0]==skipint[0]) { // delete beginning of
726 * range } else { // truncate range and create new one if necessary iv =
727 * (int[]) narange.elementAt(vc+1); if (iv[0]<=skipint[1]) { // truncate
728 * range iv[0] = skipint[1]; } else { } } } else if (iv[0]<skipint[0]) {
729 * iv = (int[]) narange.elementAt(vc+1); } } while (iv[0]) } }
731 MapList map = new MapList(scontigs, new int[]
732 { 1, resSize }, 3, 1);
734 // update newseq as if it was generated as mapping from product
738 newseq.setName(product.getSource() + "|"
739 + product.getAccessionId());
740 if (product.getMap() != null)
742 // Mapping mp = product.getMap();
743 // newseq.setStart(mp.getPosition(scontigs[0]));
745 // .getPosition(scontigs[scontigs.length - 1]));
748 transferCodedFeatures(selection, newseq, map, null, null);
749 SequenceI rseq = newseq.deriveSequence(); // construct a dataset
750 // sequence for our new
751 // peptide, regardless.
752 // store a mapping (this actually stores a mapping between the dataset
753 // sequences for the two sequences
754 codons.addMap(selection, rseq, map);
758 // register the mapping somehow
764 * Given a peptide newly translated from a dna sequence, copy over and set any
765 * features on the peptide from the DNA. If featureTypes is null, all features
766 * on the dna sequence are searched (rather than just the displayed ones), and
767 * similarly for featureGroups.
772 * @param featureTypes
773 * hash who's keys are the displayed feature type strings
774 * @param featureGroups
775 * hash where keys are feature groups and values are Boolean objects
776 * indicating if they are displayed.
778 private static void transferCodedFeatures(SequenceI dna, SequenceI pep,
779 MapList map, Hashtable featureTypes, Hashtable featureGroups)
781 SequenceFeature[] sf = (dna.getDatasetSequence() != null ? dna
782 .getDatasetSequence() : dna).getSequenceFeatures();
784 jalview.datamodel.DBRefEntry[] dnarefs = jalview.util.DBRefUtils
785 .selectRefs(dna.getDBRef(),
786 jalview.datamodel.DBRefSource.DNACODINGDBS);
789 // intersect with pep
790 for (int d = 0; d < dnarefs.length; d++)
792 Mapping mp = dnarefs[d].getMap();
800 for (int f = 0; f < sf.length; f++)
802 fgstate = (featureGroups == null) ? null : ((Boolean) featureGroups
803 .get(sf[f].featureGroup));
804 if ((featureTypes == null || featureTypes.containsKey(sf[f]
805 .getType())) && (fgstate == null || fgstate.booleanValue()))
807 if (FeatureProperties.isCodingFeature(null, sf[f].getType()))
809 // if (map.intersectsFrom(sf[f].begin, sf[f].end))