X-Git-Url: http://source.jalview.org/gitweb/?a=blobdiff_plain;f=src%2Fjalview%2Fws%2Fsifts%2FSiftsClient.java;h=53a2b8c9c20a3ff8fa1c1f75223c2b7a3693f59e;hb=c51dbe7933683b64548f5353e67561d2b5e377f7;hp=13e52cdcf92e9eb58f7e847ce5c3daac79eb9a7d;hpb=37bc56d23032f26c0da11b0532902d1f629a2080;p=jalview.git diff --git a/src/jalview/ws/sifts/SiftsClient.java b/src/jalview/ws/sifts/SiftsClient.java index 13e52cd..53a2b8c 100644 --- a/src/jalview/ws/sifts/SiftsClient.java +++ b/src/jalview/ws/sifts/SiftsClient.java @@ -21,6 +21,8 @@ package jalview.ws.sifts; import jalview.analysis.AlignSeq; +import jalview.analysis.scoremodels.ScoreMatrix; +import jalview.analysis.scoremodels.ScoreModels; import jalview.api.DBRefEntryI; import jalview.api.SiftsClientI; import jalview.datamodel.DBRefEntry; @@ -29,6 +31,9 @@ import jalview.datamodel.SequenceI; import jalview.io.StructureFile; import jalview.schemes.ResidueProperties; import jalview.structure.StructureMapping; +import jalview.structures.models.MappingOutputModel; +import jalview.util.Comparison; +import jalview.util.DBRefUtils; import jalview.util.Format; import jalview.xml.binding.sifts.Entry; import jalview.xml.binding.sifts.Entry.Entity; @@ -72,24 +77,28 @@ import MCview.PDBChain; public class SiftsClient implements SiftsClientI { + /* + * for use in mocking out file fetch for tests only + * - reset to null after testing! + */ + private static File mockSiftsFile; + + private static final int UNASSIGNED = StructureMapping.UNASSIGNED; // -1 + + private static final int PDB_RES_POS = StructureMapping.PDB_RES_NUM_INDEX; // 0 + private Entry siftsEntry; - private StructureFile pdb; + private StructureFile structureFile; private String pdbId; private String structId; - // private String segStartEnd; - private CoordinateSys seqCoordSys = CoordinateSys.UNIPROT; private static final int BUFFER_SIZE = 4096; - public static final int UNASSIGNED = -1; - - private static final int PDB_RES_POS = 0; - private static final int PDB_ATOM_POS = 1; private static final String NOT_OBSERVED = "Not_Observed"; @@ -102,7 +111,7 @@ public class SiftsClient implements SiftsClientI private HashSet curDBRefAccessionIdsString; - public enum CoordinateSys + private enum CoordinateSys { UNIPROT("UniProt"), PDB("PDBresnum"), PDBe("PDBe"); private String name; @@ -118,7 +127,7 @@ public class SiftsClient implements SiftsClientI } }; - public enum ResidueDetailType + private enum ResidueDetailType { NAME_SEC_STRUCTURE("nameSecondaryStructure"), CODE_SEC_STRUCTURE( "codeSecondaryStructure"), ANNOTATION("Annotation"); @@ -142,15 +151,14 @@ public class SiftsClient implements SiftsClientI * @param pdbId * @throws SiftsException */ - public SiftsClient(StructureFile pdb) throws SiftsException + public SiftsClient(StructureFile structureFile) throws SiftsException { - this.pdb = pdb; - this.pdbId = pdb.getId(); + this.structureFile = structureFile; + this.pdbId = structureFile.getId(); File siftsFile = getSiftsFile(pdbId); siftsEntry = parseSIFTs(siftsFile); } - /** * Parse the given SIFTs File and return a JAXB POJO of parsed data * @@ -188,6 +196,14 @@ public class SiftsClient implements SiftsClientI */ public static File getSiftsFile(String pdbId) throws SiftsException { + /* + * return mocked file if it has been set + */ + if (mockSiftsFile != null) + { + return mockSiftsFile; + } + String siftsFileName = SiftsSettings.getSiftDownloadDirectory() + pdbId.toLowerCase() + ".xml.gz"; File siftsFile = new File(siftsFileName); @@ -213,6 +229,10 @@ public class SiftsClient implements SiftsClientI return new File(siftsFileName); } } + else + { + return siftsFile; + } } try { @@ -277,21 +297,23 @@ public class SiftsClient implements SiftsClientI { siftsDownloadDir.mkdirs(); } - // System.out.println(">> Download ftp url : " + siftsFileFTPURL); - URL url = new URL(siftsFileFTPURL); - URLConnection conn = url.openConnection(); - InputStream inputStream = conn.getInputStream(); - FileOutputStream outputStream = new FileOutputStream( - downloadedSiftsFile); - byte[] buffer = new byte[BUFFER_SIZE]; - int bytesRead = -1; - while ((bytesRead = inputStream.read(buffer)) != -1) - { - outputStream.write(buffer, 0, bytesRead); - } - outputStream.close(); - inputStream.close(); - // System.out.println(">>> File downloaded : " + downloadedSiftsFile); + // System.out.println(">> Download ftp url : " + siftsFileFTPURL); + // long now = System.currentTimeMillis(); + URL url = new URL(siftsFileFTPURL); + URLConnection conn = url.openConnection(); + InputStream inputStream = conn.getInputStream(); + FileOutputStream outputStream = new FileOutputStream( + downloadedSiftsFile); + byte[] buffer = new byte[BUFFER_SIZE]; + int bytesRead = -1; + while ((bytesRead = inputStream.read(buffer)) != -1) + { + outputStream.write(buffer, 0, bytesRead); + } + outputStream.close(); + inputStream.close(); +// System.out.println(">>> File downloaded : " + downloadedSiftsFile +// + " took " + (System.currentTimeMillis() - now) + "ms"); return new File(downloadedSiftsFile); } @@ -325,41 +347,29 @@ public class SiftsClient implements SiftsClientI public DBRefEntryI getValidSourceDBRef(SequenceI seq) throws SiftsException { - DBRefEntryI sourceDBRef = null; - sourceDBRef = seq.getSourceDBRef(); - if (sourceDBRef != null && isValidDBRefEntry(sourceDBRef)) + List dbRefs = seq.getPrimaryDBRefs(); + if (dbRefs == null || dbRefs.size() < 1) { - return sourceDBRef; + throw new SiftsException( + "Source DBRef could not be determined. DBRefs might not have been retrieved."); } - else + + for (DBRefEntry dbRef : dbRefs) { - DBRefEntry[] dbRefs = seq.getDBRefs(); - if (dbRefs == null || dbRefs.length < 1) + if (dbRef == null || dbRef.getAccessionId() == null + || dbRef.getSource() == null) { - throw new SiftsException( - "Source DBRef could not be determined. DBRefs might not have been retrieved."); + continue; } - - for (DBRefEntryI dbRef : dbRefs) + String canonicalSource = DBRefUtils.getCanonicalName(dbRef + .getSource()); + if (isValidDBRefEntry(dbRef) + && (canonicalSource.equalsIgnoreCase(DBRefSource.UNIPROT) || canonicalSource + .equalsIgnoreCase(DBRefSource.PDB))) { - if (dbRef == null || dbRef.getAccessionId() == null - || dbRef.getSource() == null) - { - continue; - } - if (isFoundInSiftsEntry(dbRef.getAccessionId()) - && (dbRef.getSource().equalsIgnoreCase(DBRefSource.UNIPROT) || dbRef - .getSource().equalsIgnoreCase(DBRefSource.PDB))) - { - seq.setSourceDBRef(dbRef); - return dbRef; - } + return dbRef; } } - if (sourceDBRef != null && isValidDBRefEntry(sourceDBRef)) - { - return sourceDBRef; - } throw new SiftsException("Could not get source DB Ref"); } @@ -404,8 +414,8 @@ public class SiftsClient implements SiftsClientI String pdbFile, String chain) throws SiftsException { structId = (chain == null) ? pdbId : pdbId + "|" + chain; - System.out.println("Getting mapping for: " + pdbId + "|" + chain - + " : seq- " + seq.getName()); + System.out.println("Getting SIFTS mapping for " + structId + ": seq " + + seq.getName()); final StringBuilder mappingDetails = new StringBuilder(128); PrintStream ps = new PrintStream(System.out) @@ -442,7 +452,7 @@ public class SiftsClient implements SiftsClientI String originalSeq = AlignSeq.extractGaps( jalview.util.Comparison.GapChars, seq.getSequenceAsString()); HashMap mapping = new HashMap(); - DBRefEntryI sourceDBRef = seq.getSourceDBRef(); + DBRefEntryI sourceDBRef; sourceDBRef = getValidSourceDBRef(seq); // TODO ensure sequence start/end is in the same coordinate system and // consistent with the choosen sourceDBRef @@ -465,83 +475,9 @@ public class SiftsClient implements SiftsClientI TreeMap resNumMap = new TreeMap(); List segments = entity.getSegment(); - for (Segment segment : segments) - { - // segStartEnd = segment.getStart() + " - " + segment.getEnd(); - // System.out.println("Mapping segments : " + segment.getSegId() + "\\" - // + segStartEnd); - List residues = segment.getListResidue().getResidue(); - for (Residue residue : residues) - { - int currSeqIndex = UNASSIGNED; - List cRefDbs = residue.getCrossRefDb(); - CrossRefDb pdbRefDb = null; - for (CrossRefDb cRefDb : cRefDbs) - { - if (cRefDb.getDbSource().equalsIgnoreCase(DBRefSource.PDB)) - { - pdbRefDb = cRefDb; - } - if (cRefDb.getDbCoordSys() - .equalsIgnoreCase(seqCoordSys.getName()) - && isAccessionMatched(cRefDb.getDbAccessionId())) - { - String resNumIndexString = cRefDb.getDbResNum() - .equalsIgnoreCase("None") ? String.valueOf(UNASSIGNED) - : cRefDb.getDbResNum(); - try - { - currSeqIndex = Integer.valueOf(resNumIndexString); - } catch (NumberFormatException nfe) - { - currSeqIndex = Integer.valueOf(resNumIndexString - .split("[a-zA-Z]")[0]); - continue; - } - if (pdbRefDb != null) - { - break;// exit loop if pdb and uniprot are already found - } - } - } - if (currSeqIndex == UNASSIGNED) - { - continue; - } - if (currSeqIndex >= seq.getStart() && currSeqIndex <= seq.getEnd()) - { - int resNum; - try - { - resNum = (pdbRefDb == null) ? Integer.valueOf(residue - .getDbResNum()) : Integer.valueOf(pdbRefDb - .getDbResNum()); - } catch (NumberFormatException nfe) - { - resNum = (pdbRefDb == null) ? Integer.valueOf(residue - .getDbResNum()) : Integer.valueOf(pdbRefDb - .getDbResNum().split("[a-zA-Z]")[0]); - continue; - } - - if (isResidueObserved(residue) - || seqCoordSys == CoordinateSys.UNIPROT) - { - char resCharCode = ResidueProperties - .getSingleCharacterCode(ResidueProperties - .getCanonicalAminoAcid(residue.getDbResName())); - resNumMap.put(currSeqIndex, String.valueOf(resCharCode)); - } - else - { - omitNonObserved.add(currSeqIndex); - ++nonObservedShiftIndex; - } - mapping.put(currSeqIndex - nonObservedShiftIndex, new int[] { - Integer.valueOf(resNum), UNASSIGNED }); - } - } - } + SegmentHelperPojo shp = new SegmentHelperPojo(seq, mapping, resNumMap, + omitNonObserved, nonObservedShiftIndex); + processSegments(segments, shp); try { populateAtomPositions(entityId, mapping); @@ -558,12 +494,13 @@ public class SiftsClient implements SiftsClientI int pdbStart = UNASSIGNED; int pdbEnd = UNASSIGNED; - Integer[] keys = mapping.keySet().toArray(new Integer[0]); - Arrays.sort(keys); - if (keys.length < 1) + if (mapping.isEmpty()) { - throw new SiftsException(">>> Empty SIFTS mapping generated!!"); + throw new SiftsException("SIFTS mapping failed"); } + + Integer[] keys = mapping.keySet().toArray(new Integer[0]); + Arrays.sort(keys); seqStart = keys[0]; seqEnd = keys[keys.length - 1]; @@ -596,14 +533,14 @@ public class SiftsClient implements SiftsClientI if (os != null) { - MappingOutputPojo mop = new MappingOutputPojo(); - mop.setSeqStart(pdbStart); - mop.setSeqEnd(pdbEnd); + MappingOutputModel mop = new MappingOutputModel(); + mop.setSeqStart(seqStart); + mop.setSeqEnd(seqEnd); mop.setSeqName(seq.getName()); mop.setSeqResidue(matchedSeq); - mop.setStrStart(seqStart); - mop.setStrEnd(seqEnd); + mop.setStrStart(pdbStart); + mop.setStrEnd(pdbEnd); mop.setStrName(structId); mop.setStrResidue(targetStrucSeqs.toString()); @@ -614,6 +551,97 @@ public class SiftsClient implements SiftsClientI return mapping; } + void processSegments(List segments, SegmentHelperPojo shp) + { + SequenceI seq = shp.getSeq(); + HashMap mapping = shp.getMapping(); + TreeMap resNumMap = shp.getResNumMap(); + List omitNonObserved = shp.getOmitNonObserved(); + int nonObservedShiftIndex = shp.getNonObservedShiftIndex(); + for (Segment segment : segments) + { + // System.out.println("Mapping segments : " + segment.getSegId() + "\\"s + // + segStartEnd); + List residues = segment.getListResidue().getResidue(); + for (Residue residue : residues) + { + int currSeqIndex = UNASSIGNED; + List cRefDbs = residue.getCrossRefDb(); + CrossRefDb pdbRefDb = null; + for (CrossRefDb cRefDb : cRefDbs) + { + if (cRefDb.getDbSource().equalsIgnoreCase(DBRefSource.PDB)) + { + pdbRefDb = cRefDb; + } + if (cRefDb.getDbCoordSys() + .equalsIgnoreCase(seqCoordSys.getName()) + && isAccessionMatched(cRefDb.getDbAccessionId())) + { + currSeqIndex = getLeadingIntegerValue( + cRefDb.getDbResNum(), UNASSIGNED); + if (pdbRefDb != null) + { + break;// exit loop if pdb and uniprot are already found + } + } + } + if (currSeqIndex == UNASSIGNED) + { + continue; + } + if (currSeqIndex >= seq.getStart() && currSeqIndex <= seq.getEnd()) + { + + int resNum = (pdbRefDb == null) ? getLeadingIntegerValue( + residue.getDbResNum(), UNASSIGNED) + : getLeadingIntegerValue(pdbRefDb.getDbResNum(), + UNASSIGNED); + + if (isResidueObserved(residue) + || seqCoordSys == CoordinateSys.UNIPROT) + { + char resCharCode = ResidueProperties + .getSingleCharacterCode(ResidueProperties + .getCanonicalAminoAcid(residue.getDbResName())); + resNumMap.put(currSeqIndex, String.valueOf(resCharCode)); + } + else + { + omitNonObserved.add(currSeqIndex); + ++nonObservedShiftIndex; + } + mapping.put(currSeqIndex - nonObservedShiftIndex, new int[] { + Integer.valueOf(resNum), UNASSIGNED }); + } + } + } + } + + /** + * Get the leading integer part of a string that begins with an integer. + * + * @param input + * - the string input to process + * @param failValue + * - value returned if unsuccessful + * @return + */ + static int getLeadingIntegerValue(String input, int failValue) + { + if (input == null) + { + return failValue; + } + String[] parts = input.split("(?=\\D)(?<=\\d)"); + if (parts != null && parts.length > 0 && parts[0].matches("[0-9]+")) + { + return Integer.valueOf(parts[0]); + } + return failValue; + } + + /** * * @param chainId @@ -629,7 +657,7 @@ public class SiftsClient implements SiftsClientI { try { - PDBChain chain = pdb.findChain(chainId); + PDBChain chain = structureFile.findChain(chainId); if (chain == null || mapping == null) { @@ -751,7 +779,7 @@ public class SiftsClient implements SiftsClientI return; } Integer[] keys = resNumMap.keySet().toArray(new Integer[0]); - Arrays.sort(keys); + // Arrays.sort(keys); int firstIndex = keys[0]; int lastIndex = keys[keys.length - 1]; // System.out.println("Min value " + firstIndex); @@ -765,33 +793,11 @@ public class SiftsClient implements SiftsClientI } } - - @Override public Entity getEntityById(String id) throws SiftsException { - // Sometimes SIFTS mappings are wrongly swapped between different chains of - // a PDB entry. This results to wrong mappings being generated. The boolean - // flag 'isGetEntityIdDirectly, determines whether an entity to process is - // determined by a greedy heuristic search or by just matching the Chain Id - // directly against the entity Id tag. Setting the default value to 'false' - // utilise the heuristic search which always produces correct mappings but - // less optimised processing, where as changing the value to 'true' - // optimises performance but might result to incorrect mapping in some cases - // where SIFTS mappings are wrongly swapped between different chains. - // boolean isGetEntityIdDirectly = false; - // if (isGetEntityIdDirectly) - // { - // List entities = siftsEntry.getEntity(); - // for (Entity entity : entities) - // { - // if (!entity.getEntityId().equalsIgnoreCase(id)) - // { - // continue; - // } - // return entity; - // } - // } + // Determines an entity to process by performing a heuristic matching of all + // Entities with the given chainId and choosing the best matching Entity Entity entity = getEntityByMostOptimalMatchedId(id); if (entity != null) { @@ -840,7 +846,8 @@ public class SiftsClient implements SiftsClientI } } } - sPojo[count].pid = 100 * (sPojo[count].chainIdFreq / sPojo[count].resCount); + sPojo[count].pid = (100 * sPojo[count].chainIdFreq) + / sPojo[count].resCount; ++count; } Arrays.sort(sPojo, Collections.reverseOrder()); @@ -849,6 +856,10 @@ public class SiftsClient implements SiftsClientI if (sPojo[0].entityId != null) { + if (sPojo[0].pid < 1) + { + return null; + } for (Entity entity : entities) { if (!entity.getEntityId().equalsIgnoreCase(sPojo[0].entityId)) @@ -861,7 +872,7 @@ public class SiftsClient implements SiftsClientI return null; } - public class SiftsEntitySortPojo implements + private class SiftsEntitySortPojo implements Comparable { public String entityId; @@ -879,9 +890,83 @@ public class SiftsClient implements SiftsClientI } } + private class SegmentHelperPojo + { + private SequenceI seq; + + private HashMap mapping; + + private TreeMap resNumMap; + + private List omitNonObserved; + + private int nonObservedShiftIndex; + + public SegmentHelperPojo(SequenceI seq, + HashMap mapping, + TreeMap resNumMap, + List omitNonObserved, int nonObservedShiftIndex) + { + setSeq(seq); + setMapping(mapping); + setResNumMap(resNumMap); + setOmitNonObserved(omitNonObserved); + setNonObservedShiftIndex(nonObservedShiftIndex); + } + + public SequenceI getSeq() + { + return seq; + } + + public void setSeq(SequenceI seq) + { + this.seq = seq; + } + + public HashMap getMapping() + { + return mapping; + } + + public void setMapping(HashMap mapping) + { + this.mapping = mapping; + } + + public TreeMap getResNumMap() + { + return resNumMap; + } + + public void setResNumMap(TreeMap resNumMap) + { + this.resNumMap = resNumMap; + } + + public List getOmitNonObserved() + { + return omitNonObserved; + } + + public void setOmitNonObserved(List omitNonObserved) + { + this.omitNonObserved = omitNonObserved; + } + + public int getNonObservedShiftIndex() + { + return nonObservedShiftIndex; + } + + public void setNonObservedShiftIndex(int nonObservedShiftIndex) + { + this.nonObservedShiftIndex = nonObservedShiftIndex; + } + } @Override - public StringBuffer getMappingOutput(MappingOutputPojo mp) + public StringBuilder getMappingOutput(MappingOutputModel mp) throws SiftsException { String seqRes = mp.getSeqResidue(); @@ -903,7 +988,7 @@ public class SiftsClient implements SiftsClientI int nochunks = ((seqRes.length()) / len) + ((seqRes.length()) % len > 0 ? 1 : 0); // output mappings - StringBuffer output = new StringBuffer(); + StringBuilder output = new StringBuilder(512); output.append(NEWLINE); output.append("Sequence \u27f7 Structure mapping details").append( NEWLINE); @@ -924,6 +1009,7 @@ public class SiftsClient implements SiftsClientI output.append(String.valueOf(pdbEnd)); output.append(NEWLINE).append(NEWLINE); + ScoreMatrix pam250 = ScoreModels.getInstance().getPam250(); int matchedSeqCount = 0; for (int j = 0; j < nochunks; j++) { @@ -942,25 +1028,29 @@ public class SiftsClient implements SiftsClientI output.append(NEWLINE); output.append(new Format("%" + (maxid) + "s").form(" ")).append(" "); - // Print out the matching chars + /* + * Print out the match symbols: + * | for exact match (ignoring case) + * . if PAM250 score is positive + * else a space + */ for (int i = 0; i < len; i++) { try { if ((i + (j * len)) < seqRes.length()) { - if (seqRes.charAt(i + (j * len)) == strRes - .charAt(i + (j * len)) - && !jalview.util.Comparison.isGap(seqRes.charAt(i - + (j * len)))) + char c1 = seqRes.charAt(i + (j * len)); + char c2 = strRes.charAt(i + (j * len)); + boolean sameChar = Comparison.isSameResidue(c1, c2, false); + if (sameChar && !Comparison.isGap(c1)) { matchedSeqCount++; output.append("|"); } else if (type.equals("pep")) { - if (ResidueProperties.getPAM250(seqRes.charAt(i + (j * len)), - strRes.charAt(i + (j * len))) > 0) + if (pam250.getPairwiseScore(c1, c2) > 0) { output.append("."); } @@ -1033,4 +1123,10 @@ public class SiftsClient implements SiftsClientI return siftsEntry.getDbVersion(); } + public static void setMockSiftsFile(File file) + { + mockSiftsFile = file; + } + + }