import java.io.PrintStream;
import java.net.URL;
import java.net.URLConnection;
+import java.nio.file.Files;
+import java.nio.file.Path;
+import java.nio.file.attribute.BasicFileAttributes;
import java.util.ArrayList;
import java.util.Arrays;
import java.util.Collection;
+import java.util.Collections;
+import java.util.Date;
import java.util.HashMap;
import java.util.HashSet;
import java.util.List;
private static final String SIFTS_FTP_BASE_URL = "ftp://ftp.ebi.ac.uk/pub/databases/msd/sifts/xml/";
- public static final String DEFAULT_SIFTS_DOWNLOAD_DIR = System
- .getProperty("user.home")
- + File.separatorChar
- + ".sifts_downloads" + File.separatorChar;
-
- public static final String SIFTS_DOWNLOAD_DIR = jalview.bin.Cache
- .getDefault("sifts_download_dir", DEFAULT_SIFTS_DOWNLOAD_DIR);
-
private final static String NEWLINE = System.lineSeparator();
private String curSourceDBRef;
try (InputStream in = new FileInputStream(siftFile);
GZIPInputStream gzis = new GZIPInputStream(in);)
{
- System.out.println("File : " + siftFile.getAbsolutePath());
+ // System.out.println("File : " + siftFile.getAbsolutePath());
JAXBContext jc = JAXBContext.newInstance("jalview.xml.binding.sifts");
XMLStreamReader streamReader = XMLInputFactory.newInstance()
.createXMLStreamReader(gzis);
*/
public static File getSiftsFile(String pdbId) throws SiftsException
{
- File siftsFile = new File(SIFTS_DOWNLOAD_DIR + pdbId.toLowerCase()
- + ".xml.gz");
+ File siftsFile = new File(SiftsSettings.getSiftDownloadDirectory()
+ + pdbId.toLowerCase() + ".xml.gz");
if (siftsFile.exists())
{
- // TODO it may be worth performing an age check to determine if a
- // new SIFTs file should be re-downloaded as SIFTs entries are usually
- // updated weekly
+ // The line below is required for unit testing... don't comment it out!!!
System.out.println(">>> SIFTS File already downloaded for " + pdbId);
+
+ if (isFileOlderThanThreshold(siftsFile,
+ SiftsSettings.getCacheThresholdInDays()))
+ {
+ // System.out.println("Downloaded file is out of date, hence re-downloading...");
+ siftsFile = downloadSiftsFile(pdbId.toLowerCase());
+ }
return siftsFile;
}
siftsFile = downloadSiftsFile(pdbId.toLowerCase());
}
/**
+ * This method enables checking if a cached file has exceeded a certain
+ * threshold(in days)
+ *
+ * @param file
+ * the cached file
+ * @param noOfDays
+ * the threshold in days
+ * @return
+ */
+ public static boolean isFileOlderThanThreshold(File file, int noOfDays)
+ {
+ Path filePath = file.toPath();
+ BasicFileAttributes attr;
+ int diffInDays = 0;
+ try
+ {
+ attr = Files.readAttributes(filePath, BasicFileAttributes.class);
+ diffInDays = (int) ((new Date().getTime() - attr.lastModifiedTime()
+ .toMillis()) / (1000 * 60 * 60 * 24));
+ // System.out.println("Diff in days : " + diffInDays);
+ } catch (IOException e)
+ {
+ e.printStackTrace();
+ }
+ return noOfDays <= diffInDays;
+ }
+
+ /**
* Download a SIFTs XML file for a given PDB Id from an FTP repository
*
* @param pdbId
*/
public static File downloadSiftsFile(String pdbId) throws SiftsException
{
+ if (pdbId.contains(".cif"))
+ {
+ pdbId = pdbId.replace(".cif", "");
+ }
String siftFile = pdbId + ".xml.gz";
String siftsFileFTPURL = SIFTS_FTP_BASE_URL + siftFile;
- String downloadedSiftsFile = SIFTS_DOWNLOAD_DIR + siftFile;
- File siftsDownloadDir = new File(SIFTS_DOWNLOAD_DIR);
+ String downloadedSiftsFile = SiftsSettings.getSiftDownloadDirectory()
+ + siftFile;
+ File siftsDownloadDir = new File(
+ SiftsSettings.getSiftDownloadDirectory());
if (!siftsDownloadDir.exists())
{
siftsDownloadDir.mkdirs();
}
try
{
- System.out.println(">> Download ftp url : " + siftsFileFTPURL);
+ // System.out.println(">> Download ftp url : " + siftsFileFTPURL);
URL url = new URL(siftsFileFTPURL);
URLConnection conn = url.openConnection();
InputStream inputStream = conn.getInputStream();
}
outputStream.close();
inputStream.close();
- System.out.println(">>> File downloaded : " + downloadedSiftsFile);
+ // System.out.println(">>> File downloaded : " + downloadedSiftsFile);
} catch (IOException ex)
{
throw new SiftsException(ex.getMessage());
*/
public static boolean deleteSiftsFileByPDBId(String pdbId)
{
- File siftsFile = new File(SIFTS_DOWNLOAD_DIR + pdbId.toLowerCase()
- + ".xml.gz");
+ File siftsFile = new File(SiftsSettings.getSiftDownloadDirectory()
+ + pdbId.toLowerCase() + ".xml.gz");
if (siftsFile.exists())
{
return siftsFile.delete();
return true;
}
-
/**
* Get a valid SIFTs DBRef for the given sequence current SIFTs entry
*
DBRefEntry[] dbRefs = seq.getDBRefs();
if (dbRefs == null || dbRefs.length < 1)
{
- // final SequenceI[] seqs = new SequenceI[] { seq };
- // new jalview.ws.DBRefFetcher(seqs, null, null, null, false)
- // .fetchDBRefs(true);
- // dbRefs = seq.getDBRefs();
- }
-
- if (dbRefs == null || dbRefs.length < 1)
- {
throw new SiftsException("Could not get source DB Ref");
}
throw new SiftsException("Could not get source DB Ref");
}
-
/**
* Check that the DBRef Entry is properly populated and is available in this
* SiftClient instance
String mappingOutput = mappingDetails.toString();
StructureMapping siftsMapping = new StructureMapping(seq, pdbFile,
- pdbId, chain, mapping,
- mappingOutput);
+ pdbId, chain, mapping, mappingOutput);
return siftsMapping;
}
@Override
- public HashMap<Integer, int[]> getGreedyMapping(String entityId, SequenceI seq,
- java.io.PrintStream os)
- throws SiftsException
+ public HashMap<Integer, int[]> getGreedyMapping(String entityId,
+ SequenceI seq, java.io.PrintStream os) throws SiftsException
{
ArrayList<Integer> omitNonObserved = new ArrayList<Integer>();
int nonObservedShiftIndex = 0;
Entity entity = null;
entity = getEntityById(entityId);
String originalSeq = AlignSeq.extractGaps(
- jalview.util.Comparison.GapChars,
- seq.getSequenceAsString());
+ jalview.util.Comparison.GapChars, seq.getSequenceAsString());
HashMap<Integer, int[]> mapping = new HashMap<Integer, int[]>();
DBRefEntryI sourceDBRef = seq.getSourceDBRef();
if (sourceDBRef == null)
String resNumIndexString = cRefDb.getDbResNum()
.equalsIgnoreCase("None") ? String.valueOf(UNASSIGNED)
: cRefDb.getDbResNum();
- currSeqIndex = Integer.valueOf(resNumIndexString);
+ try
+ {
+ currSeqIndex = Integer.valueOf(resNumIndexString);
+ } catch (NumberFormatException nfe)
+ {
+ currSeqIndex = Integer.valueOf(resNumIndexString
+ .split("[a-zA-Z]")[0]);
+ }
if (pdbRefDb != null)
{
break;// exit loop if pdb and uniprot are already found
{
continue;
}
- if (currSeqIndex > seq.getStart() && currSeqIndex <= seq.getEnd())
+ if (currSeqIndex >= seq.getStart() && currSeqIndex <= seq.getEnd())
{
int resNum;
try
{
resNum = (pdbRefDb == null) ? Integer.valueOf(residue
- .getDbResNum()) : Integer.valueOf(pdbRefDb.getDbResNum());
+ .getDbResNum()) : Integer.valueOf(pdbRefDb
+ .getDbResNum());
} catch (NumberFormatException nfe)
{
resNum = (pdbRefDb == null) ? Integer.valueOf(residue
|| seqCoordSys == CoordinateSys.UNIPROT)
{
char resCharCode = ResidueProperties
- .getSingleCharacterCode(residue.getDbResName());
+ .getSingleCharacterCode(ResidueProperties
+ .getCanonicalAminoAcid(residue.getDbResName()));
resNumMap.put(currSeqIndex, String.valueOf(resCharCode));
}
else
{
e.printStackTrace();
}
- padWithGaps(resNumMap, omitNonObserved);
+ if (seqCoordSys == CoordinateSys.UNIPROT)
+ {
+ padWithGaps(resNumMap, omitNonObserved);
+ }
int seqStart = UNASSIGNED;
int seqEnd = UNASSIGNED;
int pdbStart = UNASSIGNED;
Integer[] keys = mapping.keySet().toArray(new Integer[0]);
Arrays.sort(keys);
+ if (keys.length < 1)
+ {
+ throw new SiftsException(">>> Empty SIFTS mapping generated!!");
+ }
seqStart = keys[0];
seqEnd = keys[keys.length - 1];
int orignalSeqStart = seq.getStart();
if (orignalSeqStart >= 1)
{
- int subSeqStart = seqStart - orignalSeqStart;
+ int subSeqStart = (seqStart >= orignalSeqStart) ? seqStart
+ - orignalSeqStart : 0;
int subSeqEnd = seqEnd - (orignalSeqStart - 1);
subSeqEnd = originalSeq.length() < subSeqEnd ? originalSeq.length()
: subSeqEnd;
matchedSeq = originalSeq.substring(subSeqStart, subSeqEnd);
}
+ else
+ {
+ matchedSeq = originalSeq.substring(1, originalSeq.length());
+ }
}
StringBuilder targetStrucSeqs = new StringBuilder();
if (os != null)
{
MappingOutputPojo mop = new MappingOutputPojo();
- mop.setSeqStart(seqStart);
- mop.setSeqEnd(seqEnd);
+ mop.setSeqStart(pdbStart);
+ mop.setSeqEnd(pdbEnd);
mop.setSeqName(seq.getName());
mop.setSeqResidue(matchedSeq);
- mop.setStrStart(pdbStart);
- mop.setStrEnd(pdbEnd);
+ mop.setStrStart(seqStart);
+ mop.setStrEnd(seqEnd);
mop.setStrName(structId);
mop.setStrResidue(targetStrucSeqs.toString());
}
/**
+ *
+ * @param chainId
+ * Target chain to populate mapping of its atom positions.
+ * @param mapping
+ * Two dimension array of residue index versus atom position
+ * @throws IllegalArgumentException
+ * Thrown if chainId or mapping is null
+ */
+ void populateAtomPositions(String chainId,
+ HashMap<Integer, int[]> mapping) throws IllegalArgumentException
+ {
+ PDBChain chain = pdb.findChain(chainId);
+ if (chain == null || mapping == null)
+ {
+ throw new IllegalArgumentException(
+ "Chain id or mapping must not be null.");
+ }
+ for (int[] map : mapping.values())
+ {
+ if (map[PDB_RES_POS] != UNASSIGNED)
+ {
+ map[PDB_ATOM_POS] = getAtomIndex(map[PDB_RES_POS], chain.atoms);
+ }
+ }
+ }
+
+ /**
+ *
+ * @param residueIndex
+ * The residue index used for the search
+ * @param atoms
+ * A collection of Atom to search
+ * @return atom position for the given residue index
+ */
+ int getAtomIndex(int residueIndex, Collection<Atom> atoms)
+ {
+ if (atoms == null)
+ {
+ throw new IllegalArgumentException(
+ "atoms collection must not be null!");
+ }
+ for (Atom atom : atoms)
+ {
+ if (atom.resNumber == residueIndex)
+ {
+ return atom.atomIndex;
+ }
+ }
+ return UNASSIGNED;
+ }
+
+ /**
* Checks if the residue instance is marked 'Not_observed' or not
*
* @param residue
*/
private boolean isResidueObserved(Residue residue)
{
- String annotation = getResidueAnnotaiton(residue,
+ HashSet<String> annotations = getResidueAnnotaitons(residue,
ResidueDetailType.ANNOTATION);
- if (annotation == null)
+ if (annotations == null || annotations.isEmpty())
{
return true;
}
- if (!annotation.equalsIgnoreCase(NOT_FOUND)
- && annotation.equalsIgnoreCase(NOT_OBSERVED))
+ for (String annotation : annotations)
{
- return false;
+ if (annotation.equalsIgnoreCase(NOT_OBSERVED))
+ {
+ return false;
+ }
}
return true;
}
* @param type
* @return
*/
- private String getResidueAnnotaiton(Residue residue,
+ private HashSet<String> getResidueAnnotaitons(Residue residue,
ResidueDetailType type)
{
+ HashSet<String> foundAnnotations = new HashSet<String>();
List<ResidueDetail> resDetails = residue.getResidueDetail();
for (ResidueDetail resDetail : resDetails)
{
if (resDetail.getProperty().equalsIgnoreCase(type.getCode()))
{
- return resDetail.getContent();
+ foundAnnotations.add(resDetail.getContent());
}
}
- return NOT_FOUND;
+ return foundAnnotations;
}
@Override
}
- /**
- *
- * @param chainId
- * Target chain to populate mapping of its atom positions.
- * @param mapping
- * Two dimension array of residue index versus atom position
- * @throws IllegalArgumentException
- * Thrown if chainId or mapping is null
- */
- void populateAtomPositions(String chainId, HashMap<Integer, int[]> mapping)
- throws IllegalArgumentException
+
+ @Override
+ public Entity getEntityById(String id) throws SiftsException
{
- PDBChain chain = pdb.findChain(chainId);
- if (chain == null || mapping == null)
+ // 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)
{
- throw new IllegalArgumentException(
- "Chain id or mapping must not be null.");
- }
- for (int[] map : mapping.values())
- {
- if (map[PDB_RES_POS] != UNASSIGNED)
+ List<Entity> entities = siftsEntry.getEntity();
+ for (Entity entity : entities)
{
- map[PDB_ATOM_POS] = getAtomIndex(map[PDB_RES_POS], chain.atoms);
+ if (!entity.getEntityId().equalsIgnoreCase(id))
+ {
+ continue;
+ }
+ return entity;
}
}
+ Entity entity = getEntityByMostOptimalMatchedId(id);
+ if (entity != null)
+ {
+ return entity;
+ }
+ throw new SiftsException("Entity " + id + " not found");
}
/**
+ * This method was added because EntityId is NOT always equal to ChainId.
+ * Hence, it provides the logic to greedily detect the "true" Entity for a
+ * given chainId where discrepancies exist.
*
- * @param residueIndex
- * The residue index used for the search
- * @param atoms
- * A collection of Atom to search
- * @return atom position for the given residue index
+ * @param chainId
+ * @return
*/
- int getAtomIndex(int residueIndex, Collection<Atom> atoms)
+ public Entity getEntityByMostOptimalMatchedId(String chainId)
{
- if (atoms == null)
+ // System.out.println("---> advanced greedy entityId matching block entered..");
+ List<Entity> entities = siftsEntry.getEntity();
+ SiftsEntitySortPojo[] sPojo = new SiftsEntitySortPojo[entities.size()];
+ int count = 0;
+ for (Entity entity : entities)
{
- throw new IllegalArgumentException(
- "atoms collection must not be null!");
+ sPojo[count] = new SiftsEntitySortPojo();
+ sPojo[count].entityId = entity.getEntityId();
+
+ List<Segment> segments = entity.getSegment();
+ for (Segment segment : segments)
+ {
+ List<Residue> residues = segment.getListResidue().getResidue();
+ for (Residue residue : residues)
+ {
+ List<CrossRefDb> cRefDbs = residue.getCrossRefDb();
+ for (CrossRefDb cRefDb : cRefDbs)
+ {
+ if (!cRefDb.getDbSource().equalsIgnoreCase("PDB"))
+ {
+ continue;
+ }
+ ++sPojo[count].resCount;
+ if (cRefDb.getDbChainId().equalsIgnoreCase(chainId))
+ {
+ ++sPojo[count].chainIdFreq;
+ }
+ }
+ }
+ }
+ sPojo[count].pid = 100 * (sPojo[count].chainIdFreq / sPojo[count].resCount);
+ ++count;
}
- for (Atom atom : atoms)
+ Arrays.sort(sPojo, Collections.reverseOrder());
+ System.out.println("highest matched entity : " + sPojo[0].entityId);
+ System.out.println("highest matched pid : " + sPojo[0].pid);
+
+ if (sPojo[0].entityId != null)
{
- if (atom.resNumber == residueIndex)
+ for (Entity entity : entities)
{
- return atom.atomIndex;
+ if (!entity.getEntityId().equalsIgnoreCase(sPojo[0].entityId))
+ {
+ continue;
+ }
+ return entity;
}
}
- return UNASSIGNED;
+ return null;
}
- @Override
- public Entity getEntityById(String id) throws SiftsException
+ public class SiftsEntitySortPojo implements
+ Comparable<SiftsEntitySortPojo>
{
- List<Entity> entities = siftsEntry.getEntity();
- for (Entity entity : entities)
+ public String entityId;
+
+ public int chainIdFreq;
+
+ public int pid;
+
+ public int resCount;
+
+ @Override
+ public int compareTo(SiftsEntitySortPojo o)
{
- if (!entity.getEntityId().equalsIgnoreCase(id))
- {
- continue;
- }
- return entity;
+ return this.pid - o.pid;
}
- throw new SiftsException("Entity " + id + " not found");
}
@Override
String strName = mp.getStrName();
int pdbStart = mp.getStrStart();
int pdbEnd = mp.getStrEnd();
-
+
String type = mp.getType();
-
+
int maxid = (seqName.length() >= strName.length()) ? seqName.length()
: strName.length();
int len = 72 - maxid - 1;
output.append(" - ");
output.append(String.valueOf(pdbEnd));
output.append(NEWLINE).append(NEWLINE);
-
+
int matchedSeqCount = 0;
for (int j = 0; j < nochunks; j++)
{
{
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))))
+ 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))))
+ {
matchedSeqCount++;
- output.append("|");
- }
- else if (type.equals("pep"))
- {
- if (ResidueProperties.getPAM250(seqRes.charAt(i + (j * len)),
- strRes.charAt(i + (j * len))) > 0)
+ output.append("|");
+ }
+ else if (type.equals("pep"))
{
- output.append(".");
+ if (ResidueProperties.getPAM250(seqRes.charAt(i + (j * len)),
+ strRes.charAt(i + (j * len))) > 0)
+ {
+ output.append(".");
+ }
+ else
+ {
+ output.append(" ");
+ }
}
else
{
output.append(" ");
}
}
- else
- {
- output.append(" ");
- }
- }
} catch (IndexOutOfBoundsException e)
{
continue;
output.append(NEWLINE).append(NEWLINE);
}
float pid = (float) matchedSeqCount / seqRes.length() * 100;
- if (pid < 2)
+ if (pid < SiftsSettings.getFailSafePIDThreshold())
{
- throw new SiftsException("Low PID detected for SIFTs mapping...");
+ throw new SiftsException(">>> Low PID detected for SIFTs mapping...");
}
- output.append("Length of alignment = " + seqRes.length())
- .append(NEWLINE);
+ output.append("Length of alignment = " + seqRes.length()).append(
+ NEWLINE);
output.append(new Format("Percentage ID = %2.2f").form(pid));
output.append(NEWLINE);
return output;
}
-
+
@Override
public int getEntityCount()
{
{
return siftsEntry.getDbVersion();
}
+
}
git://source.jalview.org / jalview.git / blobdiff