X-Git-Url: http://source.jalview.org/gitweb/?a=blobdiff_plain;f=src%2Fjalview%2Fws%2Fsifts%2FSiftsClient.java;h=3899f040a7b76b2202b48de31d6da71025c6ba87;hb=38889ac4815d29d1fbb040ff1dbbc47fe89ea87e;hp=29f0ffea647c5e131ac9454eee384313a0248529;hpb=984b32fdd9ae57c1a03021d70556484215317bf2;p=jalview.git diff --git a/src/jalview/ws/sifts/SiftsClient.java b/src/jalview/ws/sifts/SiftsClient.java index 29f0ffe..3899f04 100644 --- a/src/jalview/ws/sifts/SiftsClient.java +++ b/src/jalview/ws/sifts/SiftsClient.java @@ -92,14 +92,24 @@ public class SiftsClient implements SiftsClientI private CoordinateSys seqCoordSys = CoordinateSys.UNIPROT; + /** + * PDB sequence position to sequence coordinate mapping as derived from SIFTS + * record for the identified SeqCoordSys Used for lift-over from sequence + * derived from PDB (with first extracted PDBRESNUM as 'start' to the sequence + * being annotated with PDB data + */ + private jalview.datamodel.Mapping seqFromPdbMapping; + private static final int BUFFER_SIZE = 4096; - public static final int UNASSIGNED = -1; + public static final int UNASSIGNED = Integer.MIN_VALUE; private static final int PDB_RES_POS = 0; private static final int PDB_ATOM_POS = 1; + private static final int PDBE_POS = 2; + private static final String NOT_OBSERVED = "Not_Observed"; private static final String SIFTS_FTP_BASE_URL = "http://ftp.ebi.ac.uk/pub/databases/msd/sifts/xml/"; @@ -440,8 +450,9 @@ public class SiftsClient implements SiftsClientI HashMap mapping = getGreedyMapping(chain, seq, ps); String mappingOutput = mappingDetails.toString(); - StructureMapping siftsMapping = new StructureMapping(seq, pdbFile, - pdbId, chain, mapping, mappingOutput); + StructureMapping siftsMapping = new StructureMapping(aseq, pdbFile, + pdbId, chain, mapping, mappingOutput, seqFromPdbMapping); + return siftsMapping; } @@ -449,8 +460,8 @@ public class SiftsClient implements SiftsClientI public HashMap getGreedyMapping(String entityId, SequenceI seq, java.io.PrintStream os) throws SiftsException { - List omitNonObserved = new ArrayList(); - int nonObservedShiftIndex = 0; + List omitNonObserved = new ArrayList<>(); + int nonObservedShiftIndex = 0,pdbeNonObserved=0; // System.out.println("Generating mappings for : " + entityId); Entity entity = null; entity = getEntityById(entityId); @@ -481,7 +492,7 @@ public class SiftsClient implements SiftsClientI TreeMap resNumMap = new TreeMap(); List segments = entity.getSegment(); SegmentHelperPojo shp = new SegmentHelperPojo(seq, mapping, resNumMap, - omitNonObserved, nonObservedShiftIndex); + omitNonObserved, nonObservedShiftIndex,pdbeNonObserved); processSegments(segments, shp); try { @@ -503,15 +514,61 @@ public class SiftsClient implements SiftsClientI { throw new SiftsException("SIFTS mapping failed"); } + // also construct a mapping object between the seq-coord sys and the PDB seq's coord sys Integer[] keys = mapping.keySet().toArray(new Integer[0]); Arrays.sort(keys); seqStart = keys[0]; seqEnd = keys[keys.length - 1]; - + List from=new ArrayList<>(),to=new ArrayList<>(); + int[]_cfrom=null,_cto=null; String matchedSeq = originalSeq; - if (seqStart != UNASSIGNED) + if (seqStart != UNASSIGNED) // fixme! seqStart can map to -1 for a pdb sequence that starts <-1 { + for (int seqps:keys) + { + int pdbpos = mapping.get(seqps)[PDBE_POS]; + if (pdbpos == UNASSIGNED) + { + // not correct - pdbpos might be -1, but leave it for now + continue; + } + if (_cfrom==null || seqps!=_cfrom[1]+1) + { + _cfrom = new int[] { seqps,seqps}; + from.add(_cfrom); + _cto = null; // discontinuity + } else { + _cfrom[1]= seqps; + } + if (_cto==null || pdbpos!=1+_cto[1]) + { + _cto = new int[] { pdbpos,pdbpos}; + to.add(_cto); + } else { + _cto[1] = pdbpos; + } + } + _cfrom = new int[from.size() * 2]; + _cto = new int[to.size() * 2]; + int p = 0; + for (int[] range : from) + { + _cfrom[p++] = range[0]; + _cfrom[p++] = range[1]; + } + ; + p = 0; + for (int[] range : to) + { + _cto[p++] = range[0]; + _cto[p++] = range[1]; + } + ; + + seqFromPdbMapping = new jalview.datamodel.Mapping(null, _cto, _cfrom, + 1, + 1); pdbStart = mapping.get(seqStart)[PDB_RES_POS]; pdbEnd = mapping.get(seqEnd)[PDB_RES_POS]; int orignalSeqStart = seq.getStart(); @@ -564,6 +621,8 @@ public class SiftsClient implements SiftsClientI TreeMap resNumMap = shp.getResNumMap(); List omitNonObserved = shp.getOmitNonObserved(); int nonObservedShiftIndex = shp.getNonObservedShiftIndex(); + int pdbeNonObservedCount = shp.getPdbeNonObserved(); + int firstPDBResNum = UNASSIGNED; for (Segment segment : segments) { // System.out.println("Mapping segments : " + segment.getSegId() + "\\"s @@ -571,6 +630,7 @@ public class SiftsClient implements SiftsClientI List residues = segment.getListResidue().getResidue(); for (Residue residue : residues) { + boolean isObserved = isResidueObserved(residue); int pdbeIndex = getLeadingIntegerValue(residue.getDbResNum(), UNASSIGNED); int currSeqIndex = UNASSIGNED; @@ -581,6 +641,19 @@ public class SiftsClient implements SiftsClientI if (cRefDb.getDbSource().equalsIgnoreCase(DBRefSource.PDB)) { pdbRefDb = cRefDb; + if (firstPDBResNum == UNASSIGNED) + { + firstPDBResNum = getLeadingIntegerValue(cRefDb.getDbResNum(), + UNASSIGNED); + } + else + { + if (isObserved) + { + // after we find the first observed residue we just increment + firstPDBResNum++; + } + } } if (cRefDb.getDbCoordSys().equalsIgnoreCase(seqCoordSys.getName()) && isAccessionMatched(cRefDb.getDbAccessionId())) @@ -593,6 +666,10 @@ public class SiftsClient implements SiftsClientI } } } + if (!isObserved) + { + ++pdbeNonObservedCount; + } if (seqCoordSys == seqCoordSys.PDB) // FIXME: is seqCoordSys ever PDBe // ??? { @@ -605,15 +682,18 @@ public class SiftsClient implements SiftsClientI { continue; } - if (!isResidueObserved(residue) - && seqCoordSys != seqCoordSys.UNIPROT) + if (!isObserved) { - // mapping to PDB or PDBe so we need to bookkeep for the non-observed - // SEQRES positions - omitNonObserved.add(currSeqIndex); - ++nonObservedShiftIndex; + if (seqCoordSys != CoordinateSys.UNIPROT) // FIXME: PDB or PDBe only + // here + { + // mapping to PDB or PDBe so we need to bookkeep for the + // non-observed + // SEQRES positions + omitNonObserved.add(currSeqIndex); + ++nonObservedShiftIndex; + } } - // if (currSeqIndex >= seq.getStart() && currSeqIndex <= seqlength) // // true // numbering @@ -638,10 +718,13 @@ public class SiftsClient implements SiftsClientI .getCanonicalAminoAcid(residue.getDbResName())); resNumMap.put(currSeqIndex, String.valueOf(resCharCode)); } - + // TODO: mapping should have SeqCoordSysNum => int[] {PDBeNum, + // PDBRESNUM, ATOMNUM } mapping.put(currSeqIndex - nonObservedShiftIndex, new int[] - { Integer.valueOf(resNum), UNASSIGNED }); + { Integer.valueOf(resNum), UNASSIGNED, + isObserved ? firstPDBResNum + : UNASSIGNED }); } } } @@ -931,17 +1014,36 @@ public class SiftsClient implements SiftsClientI private int nonObservedShiftIndex; + /** + * count of number of 'not observed' positions in the PDB record's SEQRES + * (total number of residues with coordinates == length(SEQRES) - + * pdbeNonObserved + */ + private int pdbeNonObserved; + public SegmentHelperPojo(SequenceI seq, HashMap mapping, TreeMap resNumMap, - List omitNonObserved, int nonObservedShiftIndex) + List omitNonObserved, int nonObservedShiftIndex, + int pdbeNonObserved) { setSeq(seq); setMapping(mapping); setResNumMap(resNumMap); setOmitNonObserved(omitNonObserved); setNonObservedShiftIndex(nonObservedShiftIndex); + setPdbeNonObserved(pdbeNonObserved); + + } + + public void setPdbeNonObserved(int pdbeNonObserved2) + { + this.pdbeNonObserved = pdbeNonObserved2; } + public int getPdbeNonObserved() + { + return pdbeNonObserved; + } public SequenceI getSeq() { return seq; @@ -991,6 +1093,7 @@ public class SiftsClient implements SiftsClientI { this.nonObservedShiftIndex = nonObservedShiftIndex; } + } @Override