-package jalview.analysis;\r
-\r
-import java.util.Enumeration;\r
-import java.util.Vector;\r
-import java.util.Hashtable;\r
-\r
-import jalview.datamodel.AlignedCodonFrame;\r
-import jalview.datamodel.Alignment;\r
-import jalview.datamodel.AlignmentI;\r
-import jalview.datamodel.DBRefSource;\r
-import jalview.datamodel.DBRefEntry;\r
-import jalview.datamodel.Sequence;\r
-import jalview.datamodel.SequenceI;\r
-import jalview.ws.ASequenceFetcher;\r
-import jalview.ws.SequenceFetcher;\r
-\r
-/**\r
- * Functions for cross-referencing sequence databases. user must first specify\r
- * if cross-referencing from protein or dna (set dna==true)\r
- * \r
- * @author JimP\r
- * \r
- */\r
-public class CrossRef\r
-{\r
- /**\r
- * get the DNA or protein references for a protein or dna sequence\r
- * \r
- * @param dna\r
- * @param rfs\r
- * @return\r
- */\r
- public static DBRefEntry[] findXDbRefs(boolean dna, DBRefEntry[] rfs)\r
- {\r
- if (dna)\r
- {\r
- rfs = jalview.util.DBRefUtils.selectRefs(rfs, DBRefSource.PROTEINDBS);\r
- }\r
- else\r
- {\r
- rfs = jalview.util.DBRefUtils.selectRefs(rfs,\r
- DBRefSource.DNACODINGDBS); // could attempt to find other cross refs and return here - ie PDB xrefs (not dna, not protein seq)\r
- }\r
- return rfs;\r
- }\r
-\r
-\r
- public static Hashtable classifyDbRefs(DBRefEntry[] rfs)\r
- {\r
- Hashtable classes = new Hashtable();\r
- classes.put(DBRefSource.PROTEINDBS, jalview.util.DBRefUtils.selectRefs(rfs, DBRefSource.PROTEINDBS));\r
- classes.put(DBRefSource.DNACODINGDBS, jalview.util.DBRefUtils.selectRefs(rfs,\r
- DBRefSource.DNACODINGDBS));\r
- classes.put(DBRefSource.DOMAINDBS, jalview.util.DBRefUtils.selectRefs(rfs,\r
- DBRefSource.DOMAINDBS));\r
- // classes.put(OTHER, )\r
- return classes;\r
- }\r
-\r
- /**\r
- * @param dna\r
- * true if seqs are DNA seqs\r
- * @param seqs\r
- * @return a list of sequence database cross reference source types\r
- */\r
- public static String[] findSequenceXrefTypes(boolean dna, SequenceI[] seqs)\r
- {\r
- return findSequenceXrefTypes(dna, seqs, null);\r
- }\r
- /**\r
- * Indirect references are references from other sequences from the dataset to any of the direct\r
- * DBRefEntrys on the given sequences.\r
- * @param dna\r
- * true if seqs are DNA seqs\r
- * @param seqs\r
- * @return a list of sequence database cross reference source types\r
- */\r
- public static String[] findSequenceXrefTypes(boolean dna, SequenceI[] seqs, AlignmentI dataset)\r
- {\r
- String[] dbrefs = null;\r
- Vector refs = new Vector();\r
- for (int s = 0; s < seqs.length; s++)\r
- {\r
- SequenceI dss = seqs[s];\r
- while (dss.getDatasetSequence()!=null)\r
- {\r
- dss = dss.getDatasetSequence();\r
- }\r
- DBRefEntry[] rfs = findXDbRefs(dna, dss.getDBRef());\r
- for (int r = 0; rfs != null && r < rfs.length; r++)\r
- {\r
- if (!refs.contains(rfs[r].getSource()))\r
- {\r
- refs.addElement(rfs[r].getSource());\r
- }\r
- }\r
- if (dataset!=null)\r
- {\r
- // search for references to this sequence's direct references.\r
- DBRefEntry[] lrfs = CrossRef.findXDbRefs(!dna, seqs[s].getDBRef());\r
- Vector rseqs = new Vector();\r
- CrossRef.searchDatasetXrefs(seqs[s], !dna, lrfs, dataset, rseqs, null); // don't need to specify codon frame for mapping here\r
- Enumeration lr = rseqs.elements();\r
- while (lr.hasMoreElements())\r
- {\r
- SequenceI rs = (SequenceI) lr.nextElement();\r
- DBRefEntry[] xrs = findXDbRefs(dna, rs.getDBRef());\r
- for (int r=0; rfs != null && r < rfs.length; r++)\r
- {\r
- if (!refs.contains(rfs[r].getSource()))\r
- {\r
- refs.addElement(rfs[r].getSource());\r
- }\r
- }\r
- }\r
- }\r
- }\r
- if (refs.size() > 0)\r
- {\r
- dbrefs = new String[refs.size()];\r
- refs.copyInto(dbrefs);\r
- }\r
- return dbrefs;\r
- }\r
-\r
- /*\r
- * if (dna) { if (rfs[r].hasMap()) { // most likely this is a protein cross\r
- * reference if (!refs.contains(rfs[r].getSource())) {\r
- * refs.addElement(rfs[r].getSource()); } } }\r
- */\r
- public static boolean hasCdnaMap(SequenceI[] seqs)\r
- {\r
- String[] reftypes = findSequenceXrefTypes(false, seqs);\r
- for (int s = 0; s < reftypes.length; s++)\r
- {\r
- if (reftypes.equals(DBRefSource.EMBLCDS))\r
- {\r
- return true;\r
- // no map\r
- }\r
- }\r
- return false;\r
- }\r
-\r
- public static SequenceI[] getCdnaMap(SequenceI[] seqs)\r
- {\r
- Vector cseqs = new Vector();\r
- for (int s = 0; s < seqs.length; s++)\r
- {\r
- DBRefEntry[] cdna = findXDbRefs(true, seqs[s].getDBRef());\r
- for (int c = 0; c < cdna.length; c++)\r
- {\r
- if (cdna[c].getSource().equals(DBRefSource.EMBLCDS))\r
- {\r
- // retrieve CDS dataset sequences\r
- // need global dataset sequence retriever/resolver to reuse refs\r
- // and construct Mapping entry.\r
- // insert gaps in CDS according to peptide gaps.\r
- // add gapped sequence to cseqs\r
- }\r
- }\r
- }\r
- if (cseqs.size() > 0)\r
- {\r
- SequenceI[] rsqs = new SequenceI[cseqs.size()];\r
- cseqs.copyInto(rsqs);\r
- return rsqs;\r
- }\r
- return null;\r
-\r
- }\r
-\r
- /**\r
- * \r
- * @param dna\r
- * @param seqs\r
- * @return\r
- */\r
- public static Alignment findXrefSequences(SequenceI[] seqs, boolean dna,\r
- String source)\r
- {\r
- return findXrefSequences(seqs, dna, source, null);\r
- }\r
-\r
- /**\r
- * \r
- * @param seqs\r
- * @param dna\r
- * @param source\r
- * @param dataset\r
- * alignment to search for product sequences.\r
- * @return products (as dataset sequences)\r
- */\r
- public static Alignment findXrefSequences(SequenceI[] seqs, boolean dna,\r
- String source, AlignmentI dataset)\r
- {\r
- Vector rseqs = new Vector();\r
- Alignment ral = null;\r
- AlignedCodonFrame cf=new AlignedCodonFrame(dataset.getWidth()); // nominal width\r
- for (int s = 0; s < seqs.length; s++)\r
- {\r
- SequenceI dss = seqs[s];\r
- while (dss.getDatasetSequence()!=null)\r
- {\r
- dss = dss.getDatasetSequence();\r
- }\r
- boolean found = false;\r
- DBRefEntry[] xrfs = CrossRef.findXDbRefs(dna, dss.getDBRef());\r
- if ((xrfs == null || xrfs.length == 0) && dataset!=null)\r
- {\r
- System.out.println("Attempting to find ds Xrefs refs.");\r
- DBRefEntry[] lrfs = CrossRef.findXDbRefs(!dna, seqs[s].getDBRef()); // less ambiguous would be a 'find primary dbRefEntry' method.\r
- found = CrossRef.searchDatasetXrefs(dss, !dna, lrfs, dataset, rseqs, cf);\r
- }\r
- for (int r = 0; xrfs!=null && r < xrfs.length; r++)\r
- {\r
- if (source != null && !source.equals(xrfs[r].getSource()))\r
- continue;\r
- if (xrfs[r].hasMap())\r
- {\r
- if (xrfs[r].getMap().getTo() != null)\r
- {\r
- Sequence rsq = new Sequence(xrfs[r].getMap().getTo());\r
- rseqs.addElement(rsq);\r
- if (xrfs[r].getMap().getMap().getFromRatio()!=xrfs[r].getMap().getMap().getToRatio())\r
- {\r
- // get sense of map correct for adding to product alignment.\r
- if (dna)\r
- {\r
- // map is from dna seq to a protein product\r
- cf.addMap(dss, rsq, xrfs[r].getMap().getMap());\r
- } else {\r
- // map should be from protein seq to its coding dna\r
- cf.addMap(rsq, dss, xrfs[r].getMap().getMap().getInverse());\r
- }\r
- }\r
- found = true;\r
- }\r
- }\r
- else\r
- {\r
- // do a bit more work - search for sequences with references matching\r
- // xrefs on this sequence.\r
- if (dataset != null)\r
- {\r
- found = searchDataset(dss, xrfs[r], dataset, rseqs, cf);\r
- }\r
- }\r
- }\r
- if (!found)\r
- {\r
- if (xrfs != null && xrfs.length > 0)\r
- {\r
- // Try and get the sequence reference...\r
- /*\r
- * Ideal world - we ask for a sequence fetcher implementation here if\r
- * (jalview.io.RunTimeEnvironment.getSequenceFetcher()) (\r
- */\r
- ASequenceFetcher sftch = new SequenceFetcher();\r
- SequenceI[] retrieved = null;\r
- int l = xrfs.length;\r
- for (int r = 0; r < xrfs.length; r++)\r
- {\r
- // filter out any irrelevant or irretrievable references\r
- if ((source != null && !source.equals(xrfs[r].getSource()))\r
- || !sftch.isFetchable(xrfs[r].getSource()))\r
- {\r
- l--;\r
- xrfs[r] = null;\r
- }\r
- }\r
- if (l > 0)\r
- {\r
- System.out\r
- .println("Attempting to retrieve cross referenced sequences.");\r
- DBRefEntry[] t = new DBRefEntry[l];\r
- l = 0;\r
- for (int r = 0; r < xrfs.length; r++)\r
- {\r
- if (xrfs[r] != null)\r
- t[l++] = xrfs[r];\r
- }\r
- xrfs = t;\r
- try\r
- {\r
- retrieved = sftch.getSequences(xrfs);\r
- } catch (Exception e)\r
- {\r
- System.err\r
- .println("Problem whilst retrieving cross references for Sequence : "\r
- + seqs[s].getName());\r
- e.printStackTrace();\r
- }\r
- if (retrieved != null)\r
- {\r
- for (int rs = 0; rs < retrieved.length; rs++)\r
- {\r
- rseqs.addElement(retrieved[rs]);\r
- }\r
- }\r
- }\r
- }\r
- }\r
- }\r
- if (rseqs.size() > 0)\r
- {\r
- SequenceI[] rsqs = new SequenceI[rseqs.size()];\r
- rseqs.copyInto(rsqs);\r
- ral = new Alignment(rsqs);\r
- if (cf!=null && cf.getProtMappings()!=null)\r
- {\r
- ral.addCodonFrame(cf);\r
- }\r
- }\r
- return ral;\r
- }\r
-\r
- /**\r
- * find references to lrfs in the cross-reference set of each sequence in dataset (that is not equal to sequenceI)\r
- * Identifies matching DBRefEntry based on source and accession string only - Map and Version are nulled.\r
- * @param sequenceI\r
- * @param lrfs\r
- * @param dataset\r
- * @param rseqs\r
- * @return true if matches were found.\r
- */\r
- private static boolean searchDatasetXrefs(SequenceI sequenceI, boolean dna, DBRefEntry[] lrfs, AlignmentI dataset, Vector rseqs, AlignedCodonFrame cf)\r
- {\r
- boolean found=false;\r
- if (lrfs==null)\r
- return false;\r
- for (int i=0;i<lrfs.length; i++)\r
- {\r
- DBRefEntry xref = new DBRefEntry(lrfs[i]);\r
- // add in wildcards\r
- xref.setVersion(null);\r
- xref.setMap(null);\r
- found = searchDataset(sequenceI, xref, dataset, rseqs, cf, false, dna);\r
- }\r
- return found;\r
- }\r
-\r
-\r
- /**\r
- * search a given sequence dataset for references matching cross-references to\r
- * the given sequence\r
- * \r
- * @param sequenceI\r
- * @param xrf\r
- * @param dna\r
- * @param dataset\r
- * @param rseqs\r
- * @param cf \r
- * @return true if sequences were found and added\r
- */\r
- public static boolean searchDataset(SequenceI sequenceI, DBRefEntry xrf,\r
- AlignmentI dataset, Vector rseqs, AlignedCodonFrame cf)\r
- {\r
- return searchDataset(sequenceI, xrf,\r
- dataset, rseqs, cf, true, false);\r
- }\r
- /**\r
- * TODO: generalise to different protein classifications\r
- * Search dataset for DBRefEntrys matching the given one (xrf) and add\r
- * the associated sequence to rseq.\r
- * @param sequenceI\r
- * @param xrf\r
- * @param dataset\r
- * @param rseqs\r
- * @param direct - search all references or only subset\r
- * @param dna search dna or protein xrefs (if direct=false)\r
- * @return true if relationship found and sequence added.\r
- */\r
- public static boolean searchDataset(SequenceI sequenceI, DBRefEntry xrf,\r
- AlignmentI dataset, Vector rseqs, AlignedCodonFrame cf, boolean direct, boolean dna)\r
- {\r
- boolean found = false;\r
- if (dataset==null)\r
- return false;\r
- Enumeration e = dataset.getSequences().elements();\r
- while (e.hasMoreElements())\r
- {\r
- SequenceI nxt = (SequenceI) e.nextElement();\r
- if (nxt != null)\r
- {\r
- if (nxt.getDatasetSequence() != null)\r
- {\r
- System.err\r
- .println("Implementation warning: getProducts passed a dataset alignment without dataset sequences in it!");\r
- }\r
- if (nxt != sequenceI && nxt != sequenceI.getDatasetSequence())\r
- {\r
- DBRefEntry[] poss=null, cands=null;\r
- if (direct)\r
- {\r
- cands = jalview.util.DBRefUtils.searchRefs(poss=nxt\r
- .getDBRef(), xrf);\r
- } else {\r
- cands = jalview.util.DBRefUtils.searchRefs(\r
- poss=CrossRef.findXDbRefs(dna, nxt.getDBRef()), xrf);\r
- }\r
- if (cands != null)\r
- {\r
- rseqs.addElement(nxt);\r
- boolean foundmap= cf!=null; // don't search if we aren't given a codon map object\r
- for (int r=0; foundmap && r<cands.length; r++)\r
- {\r
- if (cands[r].hasMap())\r
- {\r
- if (cands[r].getMap().getTo()!=null && cands[r].getMap().getMap().getFromRatio()!=cands[r].getMap().getMap().getToRatio())\r
- {\r
- foundmap=true;\r
- // get sense of map correct for adding to product alignment.\r
- if (dna)\r
- {\r
- // map is from dna seq to a protein product\r
- cf.addMap(sequenceI, nxt, cands[r].getMap().getMap()); \r
- } else {\r
- // map should be from protein seq to its coding dna\r
- cf.addMap(nxt, sequenceI, cands[r].getMap().getMap().getInverse());\r
- }\r
- }\r
- }\r
- }\r
- }\r
- \r
- // TODO: add mapping between sequences if necessary\r
- found = true;\r
- }\r
- }\r
- }\r
- return found;\r
- }\r
-\r
- /**\r
- * precalculate different products that can be found for seqs in dataset\r
- * and return them.\r
- * @param dna\r
- * @param seqs\r
- * @param dataset\r
- * @param fake - don't actually build lists - just get types\r
- * @return\r
- public static Object[] buildXProductsList(boolean dna, SequenceI[] seqs, AlignmentI dataset, boolean fake)\r
- {\r
- String types[] = jalview.analysis.CrossRef.findSequenceXrefTypes(\r
- dna, seqs, dataset);\r
- if (types != null)\r
- {\r
- System.out.println("Xref Types for: "+(dna ? "dna" : "prot"));\r
- for (int t = 0; t < types.length; t++)\r
- {\r
- System.out.println("Type: " + types[t]);\r
- SequenceI[] prod = \r
- jalview.analysis.CrossRef.findXrefSequences(seqs, dna, types[t]);\r
- System.out.println("Found "\r
- + ((prod == null) ? "no" : "" + prod.length)\r
- + " products");\r
- if (prod!=null)\r
- {\r
- for (int p=0; p<prod.length; p++)\r
- {\r
- System.out.println("Prod "+p+": "+prod[p].getDisplayId(true));\r
- }\r
- }\r
- }\r
-\r
- } else {\r
- System.out.println("Trying getProducts for "+al.getSequenceAt(0).getDisplayId(true));\r
- System.out.println("Search DS Xref for: "+(dna ? "dna" : "prot"));\r
- // have a bash at finding the products amongst all the retrieved sequences.\r
- SequenceI[] prod = jalview.analysis.CrossRef.findXrefSequences(al\r
- .getSequencesArray(), dna, null, ds);\r
- System.out.println("Found "\r
- + ((prod == null) ? "no" : "" + prod.length)\r
- + " products");\r
- if (prod!=null)\r
- {\r
- // select non-equivalent sequences from dataset list\r
- for (int p=0; p<prod.length; p++)\r
- {\r
- System.out.println("Prod "+p+": "+prod[p].getDisplayId(true));\r
- }\r
- }\r
-\r
- }\r
- }\r
- */\r
-}
\ No newline at end of file
+/*
+ * Jalview - A Sequence Alignment Editor and Viewer ($$Version-Rel$$)
+ * Copyright (C) $$Year-Rel$$ The Jalview Authors
+ *
+ * This file is part of Jalview.
+ *
+ * Jalview is free software: you can redistribute it and/or
+ * modify it under the terms of the GNU General Public License
+ * as published by the Free Software Foundation, either version 3
+ * of the License, or (at your option) any later version.
+ *
+ * Jalview is distributed in the hope that it will be useful, but
+ * WITHOUT ANY WARRANTY; without even the implied warranty
+ * of MERCHANTABILITY or FITNESS FOR A PARTICULAR
+ * PURPOSE. See the GNU General Public License for more details.
+ *
+ * You should have received a copy of the GNU General Public License
+ * along with Jalview. If not, see <http://www.gnu.org/licenses/>.
+ * The Jalview Authors are detailed in the 'AUTHORS' file.
+ */
+package jalview.analysis;
+
+import jalview.datamodel.AlignedCodonFrame;
+import jalview.datamodel.Alignment;
+import jalview.datamodel.AlignmentI;
+import jalview.datamodel.DBRefEntry;
+import jalview.datamodel.DBRefSource;
+import jalview.datamodel.Mapping;
+import jalview.datamodel.Sequence;
+import jalview.datamodel.SequenceFeature;
+import jalview.datamodel.SequenceI;
+import jalview.io.gff.SequenceOntology;
+import jalview.schemes.ResidueProperties;
+import jalview.util.DBRefUtils;
+import jalview.util.MapList;
+import jalview.util.MappingUtils;
+import jalview.util.StringUtils;
+import jalview.ws.SequenceFetcher;
+import jalview.ws.seqfetcher.ASequenceFetcher;
+
+import java.util.ArrayList;
+import java.util.Collections;
+import java.util.LinkedHashMap;
+import java.util.List;
+import java.util.Map.Entry;
+import java.util.Vector;
+
+/**
+ * Functions for cross-referencing sequence databases. user must first specify
+ * if cross-referencing from protein or dna (set dna==true)
+ *
+ * @author JimP
+ *
+ */
+public class CrossRef
+{
+ /**
+ * Select just the DNA or protein references for a protein or dna sequence
+ *
+ * @param fromDna
+ * if true, select references from DNA (i.e. Protein databases), else
+ * DNA database references
+ * @param refs
+ * a set of references to select from
+ * @return
+ */
+ public static DBRefEntry[] findXDbRefs(boolean fromDna, DBRefEntry[] refs)
+ {
+ return DBRefUtils.selectRefs(refs, fromDna ? DBRefSource.PROTEINDBS
+ : DBRefSource.DNACODINGDBS);
+ // could attempt to find other cross
+ // refs here - ie PDB xrefs
+ // (not dna, not protein seq)
+ }
+
+ /**
+ * @param dna
+ * true if seqs are DNA seqs
+ * @param seqs
+ * @return a list of sequence database cross reference source types
+ */
+ public static String[] findSequenceXrefTypes(boolean dna, SequenceI[] seqs)
+ {
+ return findSequenceXrefTypes(dna, seqs, null);
+ }
+
+ /**
+ * Indirect references are references from other sequences from the dataset to
+ * any of the direct DBRefEntrys on the given sequences.
+ *
+ * @param dna
+ * true if seqs are DNA seqs
+ * @param seqs
+ * @return a list of sequence database cross reference source types
+ */
+ public static String[] findSequenceXrefTypes(boolean dna,
+ SequenceI[] seqs, AlignmentI dataset)
+ {
+ String[] dbrefs = null;
+ List<String> refs = new ArrayList<String>();
+ for (SequenceI seq : seqs)
+ {
+ if (seq != null)
+ {
+ SequenceI dss = seq;
+ while (dss.getDatasetSequence() != null)
+ {
+ dss = dss.getDatasetSequence();
+ }
+ DBRefEntry[] rfs = findXDbRefs(dna, dss.getDBRef());
+ if (rfs != null)
+ {
+ for (DBRefEntry ref : rfs)
+ {
+ if (!refs.contains(ref.getSource()))
+ {
+ refs.add(ref.getSource());
+ }
+ }
+ }
+ if (dataset != null)
+ {
+ // search for references to this sequence's direct references.
+ DBRefEntry[] lrfs = CrossRef.findXDbRefs(!dna, seq.getDBRef());
+ List<SequenceI> rseqs = new ArrayList<SequenceI>();
+ CrossRef.searchDatasetXrefs(seq, !dna, lrfs, dataset, rseqs,
+ null); // don't need to specify codon frame for mapping here
+ for (SequenceI rs : rseqs)
+ {
+ DBRefEntry[] xrs = findXDbRefs(dna, rs.getDBRef());
+ if (xrs != null)
+ {
+ for (DBRefEntry ref : xrs)
+ {
+ if (!refs.contains(ref.getSource()))
+ {
+ refs.add(ref.getSource());
+ }
+ }
+ }
+ // looks like copy and paste - change rfs to xrs?
+ // for (int r = 0; rfs != null && r < rfs.length; r++)
+ // {
+ // if (!refs.contains(rfs[r].getSource()))
+ // {
+ // refs.add(rfs[r].getSource());
+ // }
+ // }
+ }
+ }
+ }
+ }
+ if (refs.size() > 0)
+ {
+ dbrefs = new String[refs.size()];
+ refs.toArray(dbrefs);
+ }
+ return dbrefs;
+ }
+
+ public static boolean hasCdnaMap(SequenceI[] seqs)
+ {
+ // TODO unused - remove?
+ String[] reftypes = findSequenceXrefTypes(false, seqs);
+ for (int s = 0; s < reftypes.length; s++)
+ {
+ if (reftypes.equals(DBRefSource.EMBLCDS))
+ {
+ return true;
+ // no map
+ }
+ }
+ return false;
+ }
+
+ public static SequenceI[] getCdnaMap(SequenceI[] seqs)
+ {
+ // TODO unused - remove?
+ Vector cseqs = new Vector();
+ for (int s = 0; s < seqs.length; s++)
+ {
+ DBRefEntry[] cdna = findXDbRefs(true, seqs[s].getDBRef());
+ for (int c = 0; c < cdna.length; c++)
+ {
+ if (cdna[c].getSource().equals(DBRefSource.EMBLCDS))
+ {
+ System.err
+ .println("TODO: unimplemented sequence retrieval for coding region sequence.");
+ // TODO: retrieve CDS dataset sequences
+ // need global dataset sequence retriever/resolver to reuse refs
+ // and construct Mapping entry.
+ // insert gaps in CDS according to peptide gaps.
+ // add gapped sequence to cseqs
+ }
+ }
+ }
+ if (cseqs.size() > 0)
+ {
+ SequenceI[] rsqs = new SequenceI[cseqs.size()];
+ cseqs.copyInto(rsqs);
+ return rsqs;
+ }
+ return null;
+
+ }
+
+ /**
+ *
+ * @param dna
+ * @param seqs
+ * @return
+ */
+ public static Alignment findXrefSequences(SequenceI[] seqs, boolean dna,
+ String source)
+ {
+ return findXrefSequences(seqs, dna, source, null);
+ }
+
+ /**
+ *
+ * @param seqs
+ * sequences whose xrefs are being retrieved
+ * @param dna
+ * true if sequences are nucleotide
+ * @param source
+ * @param dataset
+ * alignment to search for product sequences.
+ * @return products (as dataset sequences)
+ */
+ public static Alignment findXrefSequences(SequenceI[] seqs, boolean dna,
+ String source, AlignmentI dataset)
+ {
+ List<SequenceI> rseqs = new ArrayList<SequenceI>();
+ AlignedCodonFrame cf = new AlignedCodonFrame();
+ for (SequenceI seq : seqs)
+ {
+ SequenceI dss = seq;
+ while (dss.getDatasetSequence() != null)
+ {
+ dss = dss.getDatasetSequence();
+ }
+ boolean found = false;
+ DBRefEntry[] xrfs = CrossRef.findXDbRefs(dna, dss.getDBRef());
+ if ((xrfs == null || xrfs.length == 0) && dataset != null)
+ {
+ System.out.println("Attempting to find ds Xrefs refs.");
+ // FIXME should be dss not seq here?
+ DBRefEntry[] lrfs = CrossRef.findXDbRefs(!dna, seq.getDBRef());
+ // less ambiguous would be a 'find primary dbRefEntry' method.
+ // filter for desired source xref here
+ found = CrossRef.searchDatasetXrefs(dss, !dna, lrfs, dataset,
+ rseqs, cf);
+ }
+ for (int r = 0; xrfs != null && r < xrfs.length; r++)
+ {
+ DBRefEntry xref = xrfs[r];
+ if (source != null && !source.equals(xref.getSource()))
+ {
+ continue;
+ }
+ if (xref.hasMap())
+ {
+ if (xref.getMap().getTo() != null)
+ {
+ SequenceI rsq = new Sequence(xref.getMap().getTo());
+ rseqs.add(rsq);
+ if (xref.getMap().getMap().getFromRatio() != xref
+ .getMap().getMap().getToRatio())
+ {
+ // get sense of map correct for adding to product alignment.
+ if (dna)
+ {
+ // map is from dna seq to a protein product
+ cf.addMap(dss, rsq, xref.getMap().getMap());
+ }
+ else
+ {
+ // map should be from protein seq to its coding dna
+ cf.addMap(rsq, dss, xref.getMap().getMap().getInverse());
+ }
+
+ /*
+ * compute peptide variants from dna variants
+ */
+ rsq.createDatasetSequence();
+ computeProteinVariants(seq, rsq, xref.getMap().getMap());
+ }
+ found = true;
+ }
+ }
+ if (!found)
+ {
+ // do a bit more work - search for sequences with references matching
+ // xrefs on this sequence.
+ if (dataset != null)
+ {
+ found |= searchDataset(dss, xref, dataset, rseqs, cf); // ,false,!dna);
+ if (found)
+ {
+ xrfs[r] = null; // we've recovered seqs for this one.
+ }
+ }
+ }
+ }
+ if (!found)
+ {
+ if (xrfs != null && xrfs.length > 0)
+ {
+ // Try and get the sequence reference...
+ /*
+ * Ideal world - we ask for a sequence fetcher implementation here if
+ * (jalview.io.RunTimeEnvironment.getSequenceFetcher()) (
+ */
+ ASequenceFetcher sftch = new SequenceFetcher();
+ SequenceI[] retrieved = null;
+ int l = xrfs.length;
+ for (int r = 0; r < xrfs.length; r++)
+ {
+ // filter out any irrelevant or irretrievable references
+ if (xrfs[r] == null
+ || ((source != null && !source.equals(xrfs[r]
+ .getSource())) || !sftch.isFetchable(xrfs[r]
+ .getSource())))
+ {
+ l--;
+ xrfs[r] = null;
+ }
+ }
+ if (l > 0)
+ {
+ System.out
+ .println("Attempting to retrieve cross referenced sequences.");
+ DBRefEntry[] t = new DBRefEntry[l];
+ l = 0;
+ for (int r = 0; r < xrfs.length; r++)
+ {
+ if (xrfs[r] != null)
+ {
+ t[l++] = xrfs[r];
+ }
+ }
+ xrfs = t;
+ try
+ {
+ retrieved = sftch.getSequences(xrfs, !dna);
+ // problem here is we don't know which of xrfs resulted in which
+ // retrieved element
+ } catch (Exception e)
+ {
+ System.err
+ .println("Problem whilst retrieving cross references for Sequence : "
+ + seq.getName());
+ e.printStackTrace();
+ }
+ if (retrieved != null)
+ {
+ for (int rs = 0; rs < retrieved.length; rs++)
+ {
+ // TODO: examine each sequence for 'redundancy'
+ DBRefEntry[] dbr = retrieved[rs].getDBRef();
+ if (dbr != null && dbr.length > 0)
+ {
+ for (int di = 0; di < dbr.length; di++)
+ {
+ // find any entry where we should put in the sequence being
+ // cross-referenced into the map
+ Mapping map = dbr[di].getMap();
+ if (map != null)
+ {
+ if (map.getTo() != null && map.getMap() != null)
+ {
+ // should search the local dataset to find any existing
+ // candidates for To !
+ try
+ {
+ // compare ms with dss and replace with dss in mapping
+ // if map is congruent
+ SequenceI ms = map.getTo();
+ int sf = map.getMap().getToLowest();
+ int st = map.getMap().getToHighest();
+ SequenceI mappedrg = ms.getSubSequence(sf, st);
+ SequenceI loc = dss.getSubSequence(sf, st);
+ if (mappedrg.getLength() > 0
+ && mappedrg.getSequenceAsString().equals(
+ loc.getSequenceAsString()))
+ {
+ System.err
+ .println("Mapping updated for retrieved crossreference");
+ // method to update all refs of existing To on
+ // retrieved sequence with dss and merge any props
+ // on To onto dss.
+ map.setTo(dss);
+ }
+ } catch (Exception e)
+ {
+ System.err
+ .println("Exception when consolidating Mapped sequence set...");
+ e.printStackTrace(System.err);
+ }
+ }
+ }
+ }
+ }
+ retrieved[rs].updatePDBIds();
+ rseqs.add(retrieved[rs]);
+ }
+ }
+ }
+ }
+ }
+ }
+
+ Alignment ral = null;
+ if (rseqs.size() > 0)
+ {
+ SequenceI[] rsqs = new SequenceI[rseqs.size()];
+ rseqs.toArray(rsqs);
+ ral = new Alignment(rsqs);
+ if (cf != null && !cf.isEmpty())
+ {
+ ral.addCodonFrame(cf);
+ }
+ }
+ return ral;
+ }
+
+ /**
+ * find references to lrfs in the cross-reference set of each sequence in
+ * dataset (that is not equal to sequenceI) Identifies matching DBRefEntry
+ * based on source and accession string only - Map and Version are nulled.
+ *
+ * @param sequenceI
+ * @param lrfs
+ * @param dataset
+ * @param rseqs
+ * @return true if matches were found.
+ */
+ private static boolean searchDatasetXrefs(SequenceI sequenceI,
+ boolean dna, DBRefEntry[] lrfs, AlignmentI dataset,
+ List<SequenceI> rseqs, AlignedCodonFrame cf)
+ {
+ boolean found = false;
+ if (lrfs == null)
+ {
+ return false;
+ }
+ for (int i = 0; i < lrfs.length; i++)
+ {
+ DBRefEntry xref = new DBRefEntry(lrfs[i]);
+ // add in wildcards
+ xref.setVersion(null);
+ xref.setMap(null);
+ found = searchDataset(sequenceI, xref, dataset, rseqs, cf, false, dna);
+ }
+ return found;
+ }
+
+ /**
+ * search a given sequence dataset for references matching cross-references to
+ * the given sequence
+ *
+ * @param sequenceI
+ * @param xrf
+ * @param dataset
+ * @param rseqs
+ * set of unique sequences
+ * @param cf
+ * @return true if one or more unique sequences were found and added
+ */
+ public static boolean searchDataset(SequenceI sequenceI, DBRefEntry xrf,
+ AlignmentI dataset, List<SequenceI> rseqs, AlignedCodonFrame cf)
+ {
+ return searchDataset(sequenceI, xrf, dataset, rseqs, cf, true, false);
+ }
+
+ /**
+ * TODO: generalise to different protein classifications Search dataset for
+ * DBRefEntrys matching the given one (xrf) and add the associated sequence to
+ * rseq.
+ *
+ * @param sequenceI
+ * @param xrf
+ * @param dataset
+ * @param rseqs
+ * @param direct
+ * - search all references or only subset
+ * @param dna
+ * search dna or protein xrefs (if direct=false)
+ * @return true if relationship found and sequence added.
+ */
+ public static boolean searchDataset(SequenceI sequenceI, DBRefEntry xrf,
+ AlignmentI dataset, List<SequenceI> rseqs, AlignedCodonFrame cf,
+ boolean direct, boolean dna)
+ {
+ boolean found = false;
+ SequenceI[] typer = new SequenceI[1];
+ if (dataset == null)
+ {
+ return false;
+ }
+ if (dataset.getSequences() == null)
+ {
+ System.err.println("Empty dataset sequence set - NO VECTOR");
+ return false;
+ }
+ List<SequenceI> ds;
+ synchronized (ds = dataset.getSequences())
+ {
+ for (SequenceI nxt : ds)
+ {
+ if (nxt != null)
+ {
+ if (nxt.getDatasetSequence() != null)
+ {
+ System.err
+ .println("Implementation warning: getProducts passed a dataset alignment without dataset sequences in it!");
+ }
+ if (nxt != sequenceI && nxt != sequenceI.getDatasetSequence())
+ {
+ // check if this is the correct sequence type
+ {
+ typer[0] = nxt;
+ boolean isDna = jalview.util.Comparison.isNucleotide(typer);
+ if ((direct && isDna == dna) || (!direct && isDna != dna))
+ {
+ // skip this sequence because it is same molecule type
+ continue;
+ }
+ }
+
+ // look for direct or indirect references in common
+ DBRefEntry[] poss = nxt.getDBRef(), cands = null;
+ if (direct)
+ {
+ cands = jalview.util.DBRefUtils.searchRefs(poss, xrf);
+ }
+ else
+ {
+ poss = CrossRef.findXDbRefs(dna, poss); //
+ cands = jalview.util.DBRefUtils.searchRefs(poss, xrf);
+ }
+ if (cands != null)
+ {
+ if (!rseqs.contains(nxt))
+ {
+ rseqs.add(nxt);
+ boolean foundmap = cf != null;
+ // don't search if we aren't given a codon map object
+ for (int r = 0; foundmap && r < cands.length; r++)
+ {
+ if (cands[r].hasMap())
+ {
+ if (cands[r].getMap().getTo() != null
+ && cands[r].getMap().getMap().getFromRatio() != cands[r]
+ .getMap().getMap().getToRatio())
+ {
+ foundmap = true;
+ // get sense of map correct for adding to product
+ // alignment.
+ if (dna)
+ {
+ // map is from dna seq to a protein product
+ cf.addMap(sequenceI, nxt, cands[r].getMap()
+ .getMap());
+ }
+ else
+ {
+ // map should be from protein seq to its coding dna
+ cf.addMap(nxt, sequenceI, cands[r].getMap()
+ .getMap().getInverse());
+ }
+ }
+ }
+ }
+ // TODO: add mapping between sequences if necessary
+ found = true;
+ }
+ }
+
+ }
+ }
+ }
+ }
+ return found;
+ }
+
+ /**
+ * 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
+ */
+ protected static int computeProteinVariants(SequenceI dnaSeq,
+ SequenceI peptide, 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 0;
+ }
+
+ 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] - 1));
+ 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 };
+ }
+ }
+ }
+ }
+
+ /*
+ * 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())
+ {
+ Collections.sort(peptideVariants);
+ String desc = StringUtils.listToDelimitedString(peptideVariants,
+ ", ");
+ SequenceFeature sf = new SequenceFeature(
+ SequenceOntology.SEQUENCE_VARIANT, desc, peptidePos,
+ peptidePos, Float.NaN, null);
+ peptide.getDatasetSequence().addSequenceFeature(sf);
+ count++;
+ }
+ }
+ return count;
+ }
+
+ /**
+ * Returns a 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 for codon positions 1, 2, 3
+ * @param residue
+ * the current residue translation
+ * @return
+ */
+ protected 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.equals(residue))
+ {
+ result.add(peptide);
+ }
+ }
+ }
+ }
+ return result;
+ }
+
+ /**
+ * Computes a list of all peptide variants given dna variants
+ *
+ * @param dnaSeq
+ * the coding dna sequence
+ * @param codonVariants
+ * variant features for each codon position (null if no variant)
+ * @param residue
+ * the canonical protein translation
+ * @return
+ */
+ protected static List<String> computePeptideVariants(SequenceI dnaSeq,
+ SequenceFeature[] codonVariants, String residue)
+ {
+ List<String> result = new ArrayList<String>();
+ int[][] dnaVariants = new int[3][];
+ for (int i = 0; i < 3; i++)
+ {
+
+ }
+ // TODO Auto-generated method stub
+ return null;
+ }
+
+ /**
+ * precalculate different products that can be found for seqs in dataset and
+ * return them.
+ *
+ * @param dna
+ * @param seqs
+ * @param dataset
+ * @param fake
+ * - don't actually build lists - just get types
+ * @return public static Object[] buildXProductsList(boolean dna, SequenceI[]
+ * seqs, AlignmentI dataset, boolean fake) { String types[] =
+ * jalview.analysis.CrossRef.findSequenceXrefTypes( dna, seqs,
+ * dataset); if (types != null) { System.out.println("Xref Types for:
+ * "+(dna ? "dna" : "prot")); for (int t = 0; t < types.length; t++) {
+ * System.out.println("Type: " + types[t]); SequenceI[] prod =
+ * jalview.analysis.CrossRef.findXrefSequences(seqs, dna, types[t]);
+ * System.out.println("Found " + ((prod == null) ? "no" : "" +
+ * prod.length) + " products"); if (prod!=null) { for (int p=0;
+ * p<prod.length; p++) { System.out.println("Prod "+p+":
+ * "+prod[p].getDisplayId(true)); } } } } else {
+ * System.out.println("Trying getProducts for
+ * "+al.getSequenceAt(0).getDisplayId(true));
+ * System.out.println("Search DS Xref for: "+(dna ? "dna" : "prot"));
+ * // have a bash at finding the products amongst all the retrieved
+ * sequences. SequenceI[] prod =
+ * jalview.analysis.CrossRef.findXrefSequences(al
+ * .getSequencesArray(), dna, null, ds); System.out.println("Found " +
+ * ((prod == null) ? "no" : "" + prod.length) + " products"); if
+ * (prod!=null) { // select non-equivalent sequences from dataset list
+ * for (int p=0; p<prod.length; p++) { System.out.println("Prod "+p+":
+ * "+prod[p].getDisplayId(true)); } } } }
+ */
+}
git://source.jalview.org / jalview.git / blobdiff