X-Git-Url: http://source.jalview.org/gitweb/?a=blobdiff_plain;ds=sidebyside;f=src%2Fjalview%2Fanalysis%2FCrossRef.java;h=dbad53e82d58f407b524251b875f08dd86e6e1b5;hb=a83adb45bdf9554e270921b4baad94defd314b36;hp=1a56393de39cce9c3423ba8e1bdf99008f454862;hpb=3d0101179759ef157b088ea135423cd909512d9f;p=jalview.git diff --git a/src/jalview/analysis/CrossRef.java b/src/jalview/analysis/CrossRef.java index 1a56393..dbad53e 100644 --- a/src/jalview/analysis/CrossRef.java +++ b/src/jalview/analysis/CrossRef.java @@ -31,13 +31,14 @@ import jalview.datamodel.SequenceFeature; import jalview.datamodel.SequenceI; import jalview.util.DBRefUtils; import jalview.util.MapList; -import jalview.ws.SequenceFetcherFactory; -import jalview.ws.seqfetcher.ASequenceFetcher; +import jalview.ws.SequenceFetcher; import java.util.ArrayList; import java.util.Iterator; import java.util.List; +import intervalstore.api.IntervalI; + /** * Functions for cross-referencing sequence databases. * @@ -99,7 +100,7 @@ public class CrossRef */ public List findXrefSourcesForSequences(boolean dna) { - List sources = new ArrayList(); + List sources = new ArrayList<>(); for (SequenceI seq : fromSeqs) { if (seq != null) @@ -143,15 +144,15 @@ public class CrossRef /* * first find seq's xrefs (dna-to-peptide or peptide-to-dna) */ - DBRefEntry[] rfs = DBRefUtils.selectDbRefs(!fromDna, seq.getDBRefs()); + List rfs = DBRefUtils.selectDbRefs(!fromDna, seq.getDBRefs()); addXrefsToSources(rfs, sources); if (dataset != null) { /* * find sequence's direct (dna-to-dna, peptide-to-peptide) xrefs */ - DBRefEntry[] lrfs = DBRefUtils.selectDbRefs(fromDna, seq.getDBRefs()); - List foundSeqs = new ArrayList(); + List lrfs = DBRefUtils.selectDbRefs(fromDna, seq.getDBRefs()); + List foundSeqs = new ArrayList<>(); /* * find sequences in the alignment which xref one of these DBRefs @@ -164,7 +165,7 @@ public class CrossRef */ for (SequenceI rs : foundSeqs) { - DBRefEntry[] xrs = DBRefUtils.selectDbRefs(!fromDna, + List xrs = DBRefUtils.selectDbRefs(!fromDna, rs.getDBRefs()); addXrefsToSources(xrs, sources); } @@ -178,7 +179,7 @@ public class CrossRef * @param xrefs * @param sources */ - void addXrefsToSources(DBRefEntry[] xrefs, List sources) + void addXrefsToSources(List xrefs, List sources) { if (xrefs != null) { @@ -218,7 +219,7 @@ public class CrossRef public Alignment findXrefSequences(String source, boolean fromDna) { - rseqs = new ArrayList(); + rseqs = new ArrayList<>(); AlignedCodonFrame cf = new AlignedCodonFrame(); matcher = new SequenceIdMatcher(dataset.getSequences()); @@ -230,18 +231,18 @@ public class CrossRef dss = dss.getDatasetSequence(); } boolean found = false; - DBRefEntry[] xrfs = DBRefUtils.selectDbRefs(!fromDna, + List xrfs = DBRefUtils.selectDbRefs(!fromDna, dss.getDBRefs()); // ENST & ENSP comes in to both Protein and nucleotide, so we need to // filter them // out later. - if ((xrfs == null || xrfs.length == 0) && dataset != null) + if ((xrfs == null || xrfs.size() == 0) && dataset != null) { /* * found no suitable dbrefs on sequence - look for sequences in the * alignment which share a dbref with this one */ - DBRefEntry[] lrfs = DBRefUtils.selectDbRefs(fromDna, + List lrfs = DBRefUtils.selectDbRefs(fromDna, seq.getDBRefs()); /* @@ -370,7 +371,7 @@ public class CrossRef { // do a bit more work - search for sequences with references matching // xrefs on this sequence. - found = searchDataset(fromDna, dss, xref, rseqs, cf, false); + found = searchDataset(fromDna, dss, xref, rseqs, cf, false, DBRefUtils.SEARCH_MODE_FULL); } if (found) { @@ -400,9 +401,8 @@ public class CrossRef } private void retrieveCrossRef(List sourceRefs, SequenceI seq, - DBRefEntry[] xrfs, boolean fromDna, AlignedCodonFrame cf) + List xrfs, boolean fromDna, AlignedCodonFrame cf) { - ASequenceFetcher sftch = SequenceFetcherFactory.getSequenceFetcher(); SequenceI[] retrieved = null; SequenceI dss = seq.getDatasetSequence() == null ? seq : seq.getDatasetSequence(); @@ -418,7 +418,8 @@ public class CrossRef } try { - retrieved = sftch.getSequences(sourceRefs, !fromDna); + retrieved = SequenceFetcher.getInstance() + .getSequences(sourceRefs, !fromDna); } catch (Exception e) { System.err.println( @@ -430,8 +431,8 @@ public class CrossRef if (retrieved != null) { boolean addedXref = false; - List newDsSeqs = new ArrayList(), - doNotAdd = new ArrayList(); + List newDsSeqs = new ArrayList<>(), + doNotAdd = new ArrayList<>(); for (SequenceI retrievedSequence : retrieved) { @@ -483,19 +484,24 @@ public class CrossRef private void removeAlreadyRetrievedSeqs(List sourceRefs, boolean fromDna) { - DBRefEntry[] dbrSourceSet = sourceRefs.toArray(new DBRefEntry[0]); - for (SequenceI sq : dataset.getSequences()) + List dbrSourceSet = new ArrayList<>(sourceRefs); + List dsSeqs = dataset.getSequences(); + for (int ids = 0, nds = dsSeqs.size(); ids < nds; ids++) { + SequenceI sq = dsSeqs.get(ids); boolean dupeFound = false; // !fromDna means we are looking only for nucleotide sequences, not // protein if (sq.isProtein() == fromDna) { - for (DBRefEntry dbr : sq.getPrimaryDBRefs()) + List sqdbrefs = sq.getPrimaryDBRefs(); + for (int idb = 0, ndb = sqdbrefs.size(); idb < ndb; idb++) { - for (DBRefEntry found : DBRefUtils.searchRefs(dbrSourceSet, dbr)) + DBRefEntry dbr = sqdbrefs.get(idb); + List searchrefs = DBRefUtils.searchRefs(dbrSourceSet, dbr, DBRefUtils.SEARCH_MODE_FULL); + for (int isr = 0, nsr = searchrefs.size(); isr < nsr; isr++) { - sourceRefs.remove(found); + sourceRefs.remove(searchrefs.get(isr)); dupeFound = true; } } @@ -503,7 +509,8 @@ public class CrossRef if (dupeFound) { // rebuild the search array from the filtered sourceRefs list - dbrSourceSet = sourceRefs.toArray(new DBRefEntry[0]); + dbrSourceSet.clear(); + dbrSourceSet.addAll(sourceRefs); } } } @@ -526,11 +533,13 @@ public class CrossRef * sourceSequence */ boolean imported = false; - DBRefEntry[] dbr = retrievedSequence.getDBRefs(); + List dbr = retrievedSequence.getDBRefs(); if (dbr != null) { - for (DBRefEntry dbref : dbr) + for (int ib = 0, nb = dbr.size(); ib < nb; ib++) { + + DBRefEntry dbref = dbr.get(ib); SequenceI matched = findInDataset(dbref); if (matched == sourceSequence) { @@ -542,9 +551,10 @@ public class CrossRef Mapping map = dbref.getMap(); if (map != null) { - if (map.getTo() != null && map.getMap() != null) + SequenceI ms = map.getTo(); + if (ms != null && map.getMap() != null) { - if (map.getTo() == sourceSequence) + if (ms == sourceSequence) { // already called to import once, and most likely this sequence // already imported ! @@ -555,7 +565,7 @@ public class CrossRef /* * sequence is new to dataset, so save a reference so it can be added. */ - newDsSeqs.add(map.getTo()); + newDsSeqs.add(ms); continue; } @@ -567,7 +577,6 @@ public class CrossRef { // compare ms with dss and replace with dss in mapping // if map is congruent - SequenceI ms = map.getTo(); // TODO findInDataset requires exact sequence match but // 'congruent' test is only for the mapped part // maybe not a problem in practice since only ENA provide a @@ -589,7 +598,7 @@ public class CrossRef + matched.getName(); System.out.println(msg); - DBRefEntry[] toRefs = map.getTo().getDBRefs(); + List toRefs = map.getTo().getDBRefs(); if (toRefs != null) { /* @@ -619,28 +628,36 @@ public class CrossRef * duplication (e.g. same variation from two * transcripts) */ - SequenceFeature[] sfs = ms.getSequenceFeatures(); - if (sfs != null) + List sfs = ms.getFeatures() + .getAllFeatures(); + for (SequenceFeature feat : sfs) { - for (SequenceFeature feat : sfs) + /* + * make a flyweight feature object which ignores Parent + * attribute in equality test; this avoids creating many + * otherwise duplicate exon features on genomic sequence + */ + SequenceFeature newFeature = new SequenceFeature(feat) { - /* - * make a flyweight feature object which ignores Parent - * attribute in equality test; this avoids creating many - * otherwise duplicate exon features on genomic sequence - */ - SequenceFeature newFeature = new SequenceFeature(feat) + // BH 2019.08.15 We must override equalsInterval, not + // equals, because that is part of the IntervalI interface, + // and IntervalStore may need that for proper, faster + // processing. + // @Override + // public boolean equals(Object o) + // { + // return super.equals(o, true); + // } + // + @Override + public boolean equalsInterval(IntervalI sf) { - @Override - public boolean equals(Object o) - { - return super.equals(o, true); - } - }; - matched.addSequenceFeature(newFeature); - } - } + return equals((SequenceFeature) sf, true); + } + }; + matched.addSequenceFeature(newFeature); + } } cf.addMap(retrievedSequence, map.getTo(), map.getMap()); } catch (Exception e) @@ -685,7 +702,7 @@ public class CrossRef { return; } - DBRefEntry[] dbrefs = mapTo.getDBRefs(); + List dbrefs = mapTo.getDBRefs(); if (dbrefs == null) { return; @@ -742,7 +759,7 @@ public class CrossRef { // first check primary refs. List match = DBRefUtils.searchRefs( - seq.getPrimaryDBRefs().toArray(new DBRefEntry[0]), template); + seq.getPrimaryDBRefs(), template, DBRefUtils.SEARCH_MODE_FULL); if (match != null && match.size() == 1 && sameSequence(seq, dss)) { return seq; @@ -785,15 +802,15 @@ public class CrossRef { return false; } - char[] c1 = seq1.getSequence(); - char[] c2 = seq2.getSequence(); - if (c1.length != c2.length) + + if (seq1.getLength() != seq2.getLength()) { return false; } - for (int i = 0; i < c1.length; i++) + int length = seq1.getLength(); + for (int i = 0; i < length; i++) { - int diff = c1[i] - c2[i]; + int diff = seq1.getCharAt(i) - seq2.getCharAt(i); /* * same char or differ in case only ('a'-'A' == 32) */ @@ -815,7 +832,7 @@ public class CrossRef * @param retrieved * @param acf */ - void updateDbrefMappings(SequenceI mapFrom, DBRefEntry[] xrefs, + void updateDbrefMappings(SequenceI mapFrom, List xrefs, SequenceI[] retrieved, AlignedCodonFrame acf, boolean fromDna) { SequenceIdMatcher idMatcher = new SequenceIdMatcher(retrieved); @@ -949,7 +966,7 @@ public class CrossRef * @return true if matches were found. */ private boolean searchDatasetXrefs(boolean fromDna, SequenceI sequenceI, - DBRefEntry[] lrfs, List foundSeqs, + List lrfs, List foundSeqs, AlignedCodonFrame cf) { boolean found = false; @@ -957,14 +974,14 @@ public class CrossRef { return false; } - for (int i = 0; i < lrfs.length; i++) + for (int i = 0, n = lrfs.size(); i < n; i++) { - DBRefEntry xref = new DBRefEntry(lrfs[i]); - // add in wildcards - xref.setVersion(null); - xref.setMap(null); - found |= searchDataset(fromDna, sequenceI, xref, foundSeqs, cf, - false); +// DBRefEntry xref = new DBRefEntry(lrfs.get(i)); +// // add in wildcards +// xref.setVersion(null); +// xref.setMap(null); + found |= searchDataset(fromDna, sequenceI, lrfs.get(i), foundSeqs, cf, + false, DBRefUtils.SEARCH_MODE_NO_MAP_NO_VERSION); } return found; } @@ -995,11 +1012,12 @@ public class CrossRef * sequenceI or all the returned sequences (eg a genomic reference * associated with a locus and one or more transcripts) * + * @param mode SEARCH_MODE_FULL for all; SEARCH_MODE_NO_MAP_NO_VERSION optional * @return true if relationship found and sequence added. */ boolean searchDataset(boolean fromDna, SequenceI fromSeq, DBRefEntry xrf, List foundSeqs, AlignedCodonFrame mappings, - boolean direct) + boolean direct, int mode) { boolean found = false; if (dataset == null) @@ -1011,8 +1029,8 @@ public class CrossRef System.err.println("Empty dataset sequence set - NO VECTOR"); return false; } - List ds; - synchronized (ds = dataset.getSequences()) + List ds = dataset.getSequences(); + synchronized (ds) { for (SequenceI nxt : ds) { @@ -1044,13 +1062,13 @@ public class CrossRef } // look for direct or indirect references in common - DBRefEntry[] poss = nxt.getDBRefs(); + List poss = nxt.getDBRefs(); List cands = null; // todo: indirect specifies we select either direct references to nxt // that match xrf which is indirect to sequenceI, or indirect // references to nxt that match xrf which is direct to sequenceI - cands = DBRefUtils.searchRefs(poss, xrf); + cands = DBRefUtils.searchRefs(poss, xrf, mode); // else // { // poss = DBRefUtils.selectDbRefs(nxt.isProtein()!fromDna, poss);