package jalview.analysis;\r
\r
+import java.util.Enumeration;\r
import java.util.Hashtable;\r
import java.util.Vector;\r
\r
import jalview.datamodel.AlignmentI;\r
import jalview.datamodel.Annotation;\r
import jalview.datamodel.ColumnSelection;\r
+import jalview.datamodel.DBRefEntry;\r
import jalview.datamodel.FeatureProperties;\r
import jalview.datamodel.Mapping;\r
import jalview.datamodel.Sequence;\r
return 0;\r
if (cdp1[0] < cdp2[0] || cdp1[1] < cdp2[1] || cdp1[2] < cdp2[2])\r
return -1; // one base in cdp1 precedes the corresponding base in the\r
- // other codon\r
+ // other codon\r
return 1; // one base in cdp1 appears after the corresponding base in the\r
- // other codon.\r
+ // other codon.\r
}\r
\r
/**\r
* @param gapCharacter\r
* @param annotations\r
* @param aWidth\r
+ * @param dataset destination dataset for translated sequences and mappings\r
* @return\r
*/\r
public static AlignmentI CdnaTranslate(SequenceI[] selection,\r
String[] seqstring, int viscontigs[], char gapCharacter,\r
- AlignmentAnnotation[] annotations, int aWidth)\r
+ AlignmentAnnotation[] annotations, int aWidth, Alignment dataset)\r
+ {\r
+ return CdnaTranslate(selection, seqstring, null, viscontigs,\r
+ gapCharacter, annotations, aWidth, dataset);\r
+ }\r
+\r
+ /**\r
+ * \r
+ * @param selection\r
+ * @param seqstring\r
+ * @param product - array of DbRefEntry objects from which exon map in seqstring is derived\r
+ * @param viscontigs\r
+ * @param gapCharacter\r
+ * @param annotations\r
+ * @param aWidth\r
+ * @param dataset\r
+ * @return\r
+ */\r
+ public static AlignmentI CdnaTranslate(SequenceI[] selection,\r
+ String[] seqstring, DBRefEntry[] product, int viscontigs[],\r
+ char gapCharacter, AlignmentAnnotation[] annotations, int aWidth, Alignment dataset)\r
{\r
AlignedCodonFrame codons = new AlignedCodonFrame(aWidth); // stores hash of\r
- // subsequent\r
- // positions for\r
- // each codon\r
- // start position\r
- // in alignment\r
+ // subsequent\r
+ // positions for\r
+ // each codon\r
+ // start position\r
+ // in alignment\r
int s, sSize = selection.length;\r
Vector pepseqs = new Vector();\r
for (s = 0; s < sSize; s++)\r
{\r
SequenceI newseq = translateCodingRegion(selection[s], seqstring[s],\r
- viscontigs, codons, gapCharacter);\r
+ viscontigs, codons, gapCharacter, (product!=null) ? product[s] : null); // possibly anonymous product\r
if (newseq != null)\r
{\r
pepseqs.addElement(newseq);\r
+ SequenceI ds = newseq;\r
+ while (ds.getDatasetSequence()!=null)\r
+ {\r
+ ds = ds.getDatasetSequence();\r
+ }\r
+ dataset.addSequence(ds);\r
}\r
}\r
if (codons.aaWidth == 0)\r
pepseqs.copyInto(newseqs);\r
AlignmentI al = new Alignment(newseqs);\r
al.padGaps(); // ensure we look aligned.\r
- al.setDataset(null);\r
+ al.setDataset(dataset);\r
translateAlignedAnnotations(annotations, al, codons);\r
al.addCodonFrame(codons);\r
return al;\r
}\r
+ /**\r
+ * fake the collection of DbRefs with associated exon mappings to identify\r
+ * if a translation would generate distinct product in the currently selected region.\r
+ * @param selection\r
+ * @param viscontigs\r
+ * @return\r
+ */\r
+ public static boolean canTranslate(SequenceI[] selection, int viscontigs[])\r
+ {\r
+ for (int gd=0; gd<selection.length; gd++)\r
+ {\r
+ SequenceI dna = selection[gd];\r
+ jalview.datamodel.DBRefEntry[] dnarefs = jalview.util.DBRefUtils\r
+ .selectRefs(dna.getDBRef(),\r
+ jalview.datamodel.DBRefSource.DNACODINGDBS);\r
+ if (dnarefs != null)\r
+ {\r
+ // intersect with pep\r
+ // intersect with pep\r
+ Vector mappedrefs = new Vector();\r
+ DBRefEntry[] refs=dna.getDBRef();\r
+ for (int d=0; d<refs.length; d++)\r
+ {\r
+ if (refs[d].getMap()!=null && refs[d].getMap().getMap()!=null && refs[d].getMap().getMap().getFromRatio()==3\r
+ && refs[d].getMap().getMap().getToRatio()==1)\r
+ {\r
+ mappedrefs.addElement(refs[d]); // add translated protein maps\r
+ }\r
+ }\r
+ dnarefs = new DBRefEntry[mappedrefs.size()];\r
+ mappedrefs.copyInto(dnarefs);\r
+ for (int d = 0; d < dnarefs.length; d++)\r
+ {\r
+ Mapping mp = dnarefs[d].getMap();\r
+ if (mp != null)\r
+ {\r
+ for (int vc=0; vc<viscontigs.length; vc+=2)\r
+ {\r
+ int[] mpr=mp.locateMappedRange(viscontigs[vc], viscontigs[vc+1]);\r
+ if (mpr!=null) {\r
+ return true;\r
+ }\r
+ }\r
+ }\r
+ }\r
+ }\r
+ }\r
+ return false;\r
+ }\r
+ /**\r
+ * generate a set of translated protein products from annotated sequenceI\r
+ * @param selection\r
+ * @param viscontigs\r
+ * @param gapCharacter\r
+ * @param dataset destination dataset for translated sequences\r
+ * @param annotations\r
+ * @param aWidth\r
+ * @return\r
+ */\r
+ public static AlignmentI CdnaTranslate(SequenceI[] selection,\r
+ int viscontigs[], char gapCharacter, Alignment dataset)\r
+ {\r
+ int alwidth=0;\r
+ Vector cdnasqs=new Vector();\r
+ Vector cdnasqi=new Vector();\r
+ Vector cdnaprod=new Vector();\r
+ for (int gd=0; gd<selection.length; gd++)\r
+ {\r
+ SequenceI dna = selection[gd];\r
+ jalview.datamodel.DBRefEntry[] dnarefs = jalview.util.DBRefUtils\r
+ .selectRefs(dna.getDBRef(),\r
+ jalview.datamodel.DBRefSource.DNACODINGDBS);\r
+ if (dnarefs != null)\r
+ {\r
+ // intersect with pep\r
+ Vector mappedrefs = new Vector();\r
+ DBRefEntry[] refs=dna.getDBRef();\r
+ for (int d=0; d<refs.length; d++)\r
+ {\r
+ if (refs[d].getMap()!=null && refs[d].getMap().getMap()!=null && refs[d].getMap().getMap().getFromRatio()==3\r
+ && refs[d].getMap().getMap().getToRatio()==1)\r
+ {\r
+ mappedrefs.addElement(refs[d]); // add translated protein maps\r
+ }\r
+ }\r
+ dnarefs = new DBRefEntry[mappedrefs.size()];\r
+ mappedrefs.copyInto(dnarefs);\r
+ for (int d = 0; d < dnarefs.length; d++)\r
+ {\r
+ Mapping mp = dnarefs[d].getMap();\r
+ StringBuffer sqstr=new StringBuffer();\r
+ if (mp != null)\r
+ {\r
+ Mapping intersect = mp.intersectVisContigs(viscontigs);\r
+ // generate seqstring for this sequence based on mapping\r
+ \r
+ if (sqstr.length()>alwidth)\r
+ alwidth = sqstr.length();\r
+ cdnasqs.addElement(sqstr.toString());\r
+ cdnasqi.addElement(dna);\r
+ cdnaprod.addElement(intersect);\r
+ }\r
+ }\r
+ }\r
+ SequenceI[] cdna = new SequenceI[cdnasqs.size()];\r
+ DBRefEntry[] prods = new DBRefEntry[cdnaprod.size()];\r
+ String[] xons = new String[cdnasqs.size()];\r
+ cdnasqs.copyInto(xons);\r
+ cdnaprod.copyInto(prods);\r
+ cdnasqi.copyInto(cdna);\r
+ return CdnaTranslate(cdna, xons, prods, viscontigs, gapCharacter, null, alwidth, dataset);\r
+ }\r
+ return null;\r
+ }\r
\r
/**\r
* translate na alignment annotations onto translated amino acid alignment al\r
if (codons.codons[a] != null\r
&& codons.codons[a][0] == (codons.codons[a][2] - 2))\r
{\r
- pos = codons.codons[a][0];\r
- if (annotations[i].annotations[pos] == null\r
- || annotations[i].annotations[pos] == null)\r
- continue;\r
- // We just take the annotation in the first base in the codon\r
- anots[a] = new Annotation(annotations[i].annotations[pos]);\r
+ anots[a] = getCodonAnnotation(codons.codons[a], annotations[i].annotations);\r
}\r
}\r
}\r
\r
jalview.datamodel.AlignmentAnnotation aa = new jalview.datamodel.AlignmentAnnotation(\r
annotations[i].label, annotations[i].description, anots);\r
+ aa.graph = annotations[i].graph;\r
+ aa.graphGroup = annotations[i].graphGroup;\r
+ aa.graphHeight = annotations[i].graphHeight;\r
+ if (annotations[i].getThreshold()!=null)\r
+ {\r
+ aa.setThreshold(new jalview.datamodel.GraphLine(annotations[i].getThreshold())); \r
+ }\r
if (annotations[i].hasScore)\r
{\r
aa.setScore(annotations[i].getScore());\r
// positioning\r
aa.setSequenceRef(aaSeq);\r
aa.createSequenceMapping(aaSeq, aaSeq.getStart(), true); // rebuild\r
- // mapping\r
+ // mapping\r
aa.adjustForAlignment();\r
aaSeq.addAlignmentAnnotation(aa);\r
}\r
}\r
}\r
\r
+ private static Annotation getCodonAnnotation(int[] is, Annotation[] annotations)\r
+ { \r
+ // Have a look at all the codon positions for annotation and put the first\r
+ // one found into the translated annotation pos.\r
+ for (int p=0; p<3; p++)\r
+ {\r
+ if (annotations[is[p]]!=null)\r
+ {\r
+ return new Annotation(annotations[is[p]]);\r
+ }\r
+ }\r
+ return null;\r
+ }\r
+\r
/**\r
* Translate a na sequence\r
* \r
- * @param selection\r
- * @param seqstring\r
- * @param viscontigs\r
- * @param codons\r
+ * @param selection sequence displayed under viscontigs visible columns\r
+ * @param seqstring ORF read in some global alignment reference frame\r
+ * @param viscontigs mapping from global reference frame to visible seqstring ORF read\r
+ * @param codons Definition of global ORF alignment reference frame\r
* @param gapCharacter\r
* @param newSeq\r
* @return sequence ready to be added to alignment.\r
*/\r
public static SequenceI translateCodingRegion(SequenceI selection,\r
String seqstring, int[] viscontigs, AlignedCodonFrame codons,\r
- char gapCharacter)\r
+ char gapCharacter, DBRefEntry product)\r
{\r
ShiftList vismapping = new ShiftList(); // map from viscontigs to seqstring\r
- // intervals\r
+ // intervals\r
int vc, scontigs[] = new int[viscontigs.length];\r
int npos = 0;\r
for (vc = 0; vc < viscontigs.length; vc += 2)\r
// with a gap\r
aa = "" + gapCharacter + aa;\r
aspos++;\r
- if (aspos >= codons.aaWidth)\r
- codons.aaWidth = aspos + 1;\r
+ //if (aspos >= codons.aaWidth)\r
+ // codons.aaWidth = aspos + 1;\r
break; // check the next position for alignment\r
case 0:\r
// codon aligns at aspos position.\r
// map and trim contigs to ORF region\r
vc = scontigs.length - 1;\r
lastnpos = vismapping.shift(lastnpos); // place npos in context of\r
- // whole dna alignment (rather\r
- // than visible contigs)\r
+ // whole dna alignment (rather\r
+ // than visible contigs)\r
// incomplete ORF could be broken over one or two visible contig\r
// intervals.\r
while (vc >= 0 && scontigs[vc] > lastnpos)\r
System.arraycopy(scontigs, 0, t, 0, vc + 2);\r
scontigs = t;\r
}\r
-\r
MapList map = new MapList(scontigs, new int[]\r
- { 1, resSize }, 3, 1); // TODO: store mapping on newSeq for linked\r
- // DNA/Protein viewing.\r
+ { 1, resSize }, 3, 1);\r
+ // update newseq as if it was generated as mapping from product\r
+ \r
+ if (product != null)\r
+ {\r
+ newseq.setName(product.getSource() + "|"\r
+ + product.getAccessionId());\r
+ if (product.getMap() != null)\r
+ {\r
+ //Mapping mp = product.getMap();\r
+ //newseq.setStart(mp.getPosition(scontigs[0]));\r
+ //newseq.setEnd(mp\r
+ // .getPosition(scontigs[scontigs.length - 1]));\r
+ }\r
+ }\r
transferCodedFeatures(selection, newseq, map, null, null);\r
SequenceI rseq = newseq.deriveSequence(); // construct a dataset\r
- // sequence for our new\r
- // peptide, regardless.\r
+ // sequence for our new\r
+ // peptide, regardless.\r
// store a mapping (this actually stores a mapping between the dataset\r
// sequences for the two sequences\r
- codons.addMap(selection, newseq, map);\r
+ codons.addMap(selection, rseq, map);\r
return rseq;\r
}\r
}\r
git://source.jalview.org / jalview.git / commitdiff