3 * Jalview - A Sequence Alignment Editor and Viewer ($$Version-Rel$$)
4 * Copyright (C) $$Year-Rel$$ The Jalview Authors
6 * This file is part of Jalview.
8 * Jalview is free software: you can redistribute it and/or
9 * modify it under the terms of the GNU General Public License
10 * as published by the Free Software Foundation, either version 3
11 * of the License, or (at your option) any later version.
13 * Jalview is distributed in the hope that it will be useful, but
14 * WITHOUT ANY WARRANTY; without even the implied warranty
15 * of MERCHANTABILITY or FITNESS FOR A PARTICULAR
16 * PURPOSE. See the GNU General Public License for more details.
18 * You should have received a copy of the GNU General Public License
19 * along with Jalview. If not, see <http://www.gnu.org/licenses/>.
20 * The Jalview Authors are detailed in the 'AUTHORS' file.
22 package jalview.ws.dbsources;
25 import java.io.FileInputStream;
26 import java.io.FileNotFoundException;
27 import java.io.IOException;
28 import java.io.InputStream;
29 import java.util.ArrayList;
30 import java.util.Date;
31 import java.util.HashMap;
32 import java.util.List;
35 import org.json.simple.JSONArray;
36 import org.json.simple.JSONObject;
37 import org.json.simple.parser.ParseException;
39 import com.stevesoft.pat.Regex;
41 import jalview.api.FeatureSettingsModelI;
42 import jalview.bin.Console;
43 import jalview.datamodel.AlignmentAnnotation;
44 import jalview.datamodel.AlignmentI;
45 import jalview.datamodel.ContactMatrixI;
46 import jalview.datamodel.DBRefEntry;
47 import jalview.datamodel.PDBEntry;
48 import jalview.datamodel.SequenceFeature;
49 import jalview.datamodel.SequenceI;
50 import jalview.gui.Desktop;
51 import jalview.io.DataSourceType;
52 import jalview.io.FileFormat;
53 import jalview.io.FileFormatI;
54 import jalview.io.FormatAdapter;
55 import jalview.io.PDBFeatureSettings;
56 import jalview.structure.StructureImportSettings.TFType;
57 import jalview.structure.StructureMapping;
58 import jalview.structure.StructureSelectionManager;
59 import jalview.util.MessageManager;
60 import jalview.util.Platform;
61 import jalview.ws.datamodel.alphafold.PAEContactMatrix;
62 import jalview.ws.utils.UrlDownloadClient;
68 public class EBIAlfaFold extends EbiFileRetrievedProxy
70 private static final String SEPARATOR = "|";
72 private static final String COLON = ":";
74 private static final int PDB_ID_LENGTH = 4;
76 private static String AF_VERSION = "3";
86 * @see jalview.ws.DbSourceProxy#getAccessionSeparator()
89 public String getAccessionSeparator()
97 * @see jalview.ws.DbSourceProxy#getAccessionValidator()
100 public Regex getAccessionValidator()
102 Regex validator = new Regex("(AF-[A-Z]+[0-9]+[A-Z0-9]+-F1)");
103 validator.setIgnoreCase(true);
110 * @see jalview.ws.DbSourceProxy#getDbSource()
113 public String getDbSource()
121 * @see jalview.ws.DbSourceProxy#getDbVersion()
124 public String getDbVersion()
129 public static String getAlphaFoldCifDownloadUrl(String id, String vnum)
131 if (vnum == null || vnum.length() == 0)
135 return "https://alphafold.ebi.ac.uk/files/" + id + "-model_v" + vnum
139 public static String getAlphaFoldPaeDownloadUrl(String id, String vnum)
141 if (vnum == null || vnum.length() == 0)
145 return "https://alphafold.ebi.ac.uk/files/" + id
146 + "-predicted_aligned_error_v" + vnum + ".json";
152 * @see jalview.ws.DbSourceProxy#getSequenceRecords(java.lang.String[])
155 public AlignmentI getSequenceRecords(String queries) throws Exception
157 // used by the SequenceFetcher. Put forSequencesOnly to true (don't open
158 // structure viewer later)
159 return getSequenceRecords(queries, null, true);
162 public AlignmentI getSequenceRecords(String queries, String retrievalUrl)
165 return getSequenceRecords(queries, retrievalUrl, false);
168 public AlignmentI getSequenceRecords(String queries, String retrievalUrl,
169 boolean forSequencesOnly) throws Exception
171 AlignmentI pdbAlignment = null;
174 if (queries.indexOf(COLON) > -1)
176 chain = queries.substring(queries.indexOf(COLON) + 1);
177 id = queries.substring(0, queries.indexOf(COLON));
184 if (!isValidReference(id))
187 "(AFClient) Ignoring invalid alphafold query: '" + id + "'");
191 String alphaFoldCif = getAlphaFoldCifDownloadUrl(id, AF_VERSION);
192 if (retrievalUrl != null)
194 alphaFoldCif = retrievalUrl;
199 File tmpFile = File.createTempFile(id, ".cif");
200 Console.debug("Retrieving structure file for " + id + " from "
202 UrlDownloadClient.download(alphaFoldCif, tmpFile);
204 // may not need this check ?
205 file = tmpFile.getAbsolutePath();
210 // TODO Get the PAE file somewhere around here and remove from JmolParser
212 pdbAlignment = importDownloadedStructureFromUrl(alphaFoldCif, tmpFile,
213 id, chain, getDbSource(), getDbVersion());
215 if (pdbAlignment == null || pdbAlignment.getHeight() < 1)
217 throw new Exception(MessageManager.formatMessage(
218 "exception.no_pdb_records_for_chain", new String[]
219 { id, ((chain == null) ? "' '" : chain) }));
221 // done during structure retrieval
222 // retrieve_AlphaFold_pAE(id, pdbAlignment, retrievalUrl);
224 } catch (Exception ex) // Problem parsing PDB file
229 // distinguish between AlphaFold cif/pdb opened as structure file or fetched
230 // as sequence/alignment
235 * get an alphafold pAE for the given id and return the File object of the
236 * downloaded (temp) file
239 * @param pdbAlignment
240 * @param retrievalUrl
241 * - URL of .mmcif from EBI-AlphaFold - will be used to generate the
242 * pAE URL automatically
243 * @throws IOException
246 public static File fetchAlphaFoldPAE(String id, String retrievalUrl)
249 // import PAE as contact matrix - assume this will work if there was a
251 String paeURL = getAlphaFoldPaeDownloadUrl(id, AF_VERSION);
253 if (retrievalUrl != null)
255 // manufacture the PAE url from a url like ...-model-vN.cif
256 paeURL = retrievalUrl.replace("model", "predicted_aligned_error")
257 .replace(".cif", ".json");
261 File pae = paeDownloadCache.get(paeURL);
262 if (pae != null && pae.exists() && (new Date().getTime()
263 - pae.lastModified()) < PAE_CACHE_STALE_TIME)
266 "Using existing file in PAE cache for '" + paeURL + "'");
272 pae = File.createTempFile(id == null ? "af_pae" : id, "pae_json");
273 } catch (IOException e)
277 Console.debug("Downloading pae from " + paeURL + " to " + pae.toString()
281 UrlDownloadClient.download(paeURL, pae);
282 } catch (IOException e)
286 // cache and it if successful
287 paeDownloadCache.put(paeURL, pae);
292 * get an alphafold pAE for the given id, and add it to sequence 0 in
293 * pdbAlignment (assuming it came from structurefile parser).
296 * @param pdbAlignment
297 * @param retrievalUrl
298 * - URL of .mmcif from EBI-AlphaFold - will be used to generate the
299 * pAE URL automatically
300 * @throws IOException
303 public static void retrieve_AlphaFold_pAE(String id,
304 AlignmentI pdbAlignment, String retrievalUrl) throws IOException
306 File pae = fetchAlphaFoldPAE(id, retrievalUrl);
307 addAlphaFoldPAE(pdbAlignment, pae, 0, null, false, false);
310 public static void addAlphaFoldPAE(AlignmentI pdbAlignment, File pae,
311 int index, String id, boolean isStruct, boolean isStructId)
313 FileInputStream paeInput = null;
316 paeInput = new FileInputStream(pae);
317 } catch (FileNotFoundException e)
320 "Could not find pAE file '" + pae.getAbsolutePath() + "'", e);
326 StructureSelectionManager ssm = StructureSelectionManager
327 .getStructureSelectionManager(Desktop.instance);
330 String structFilename = isStructId ? ssm.findFileForPDBId(id) : id;
331 addPAEToStructure(ssm, structFilename, pae);
337 // attach to sequence?!
340 if (!importPaeJSONAsContactMatrixToSequence(pdbAlignment, paeInput,
343 Console.warn("Could not import contact matrix from '"
344 + pae.getAbsolutePath() + "' to sequence.");
346 } catch (IOException e1)
348 Console.error("Error when importing pAE file '"
349 + pae.getAbsolutePath() + "'", e1);
350 } catch (ParseException e2)
352 Console.error("Error when parsing pAE file '"
353 + pae.getAbsolutePath() + "'", e2);
359 public static void addPAEToStructure(StructureSelectionManager ssm,
360 String structFilename, File pae)
362 FileInputStream paeInput = null;
365 paeInput = new FileInputStream(pae);
366 } catch (FileNotFoundException e)
369 "Could not find pAE file '" + pae.getAbsolutePath() + "'", e);
374 ssm = StructureSelectionManager
375 .getStructureSelectionManager(Desktop.instance);
379 StructureMapping[] smArray = ssm.getMapping(structFilename);
383 if (!importPaeJSONAsContactMatrixToStructure(smArray, paeInput))
385 Console.warn("Could not import contact matrix from '"
386 + pae.getAbsolutePath() + "' to structure.");
388 } catch (IOException e1)
390 Console.error("Error when importing pAE file '"
391 + pae.getAbsolutePath() + "'", e1);
392 } catch (ParseException e2)
394 Console.error("Error when parsing pAE file '"
395 + pae.getAbsolutePath() + "'", e2);
401 * parses the given pAE matrix and adds it to sequence 0 in the given
404 * @param pdbAlignment
406 * @return true if there was a pAE matrix added
407 * @throws ParseException
408 * @throws IOException
411 public static boolean importPaeJSONAsContactMatrixToSequence(
412 AlignmentI pdbAlignment, InputStream pae_input, int index,
413 String seqId) throws IOException, ParseException
415 SequenceI sequence = null;
418 int seqToGet = index > 0 ? index : 0;
419 sequence = pdbAlignment.getSequenceAt(seqToGet);
421 if (sequence == null)
423 SequenceI[] sequences = pdbAlignment.findSequenceMatch(seqId);
424 if (sequences == null || sequences.length < 1)
426 Console.warn("Could not find sequence with id '" + seqId
427 + "' to attach pAE matrix to. Ignoring matrix.");
432 sequence = sequences[0]; // just use the first sequence with this seqId
435 if (sequence == null)
439 return importPaeJSONAsContactMatrixToSequence(pdbAlignment, pae_input,
443 public static boolean importPaeJSONAsContactMatrixToSequence(
444 AlignmentI pdbAlignment, InputStream pae_input,
445 SequenceI sequence) throws IOException, ParseException
447 JSONObject paeDict = parseJSONtoPAEContactMatrix(pae_input);
450 Console.debug("JSON file did not parse properly.");
453 ContactMatrixI matrix = new PAEContactMatrix(sequence,
454 (Map<String, Object>) paeDict);
456 AlignmentAnnotation cmannot = sequence.addContactList(matrix);
457 pdbAlignment.addAnnotation(cmannot);
462 public static JSONObject parseJSONtoPAEContactMatrix(
463 InputStream pae_input) throws IOException, ParseException
465 Object paeJson = Platform.parseJSON(pae_input);
466 JSONObject paeDict = null;
467 if (paeJson instanceof JSONObject)
469 paeDict = (JSONObject) paeJson;
471 else if (paeJson instanceof JSONArray)
473 JSONArray jsonArray = (JSONArray) paeJson;
474 if (jsonArray.size() > 0)
475 paeDict = (JSONObject) jsonArray.get(0);
481 public static boolean importPaeJSONAsContactMatrixToStructure(
482 StructureMapping[] smArray, InputStream paeInput)
483 throws IOException, ParseException
485 boolean someDone = false;
486 for (StructureMapping sm : smArray)
488 boolean thisDone = importPaeJSONAsContactMatrixToStructure(sm,
490 someDone |= thisDone;
495 public static boolean importPaeJSONAsContactMatrixToStructure(
496 StructureMapping sm, InputStream paeInput)
497 throws IOException, ParseException
499 JSONObject pae_obj = parseJSONtoPAEContactMatrix(paeInput);
502 Console.debug("JSON file did not parse properly.");
506 ContactMatrixI matrix = new PAEContactMatrix(sm.getSequence(),
507 (Map<String, Object>) pae_obj);
509 AlignmentAnnotation cmannot = sm.getSequence().addContactList(matrix);
510 sm.getSequence().addAlignmentAnnotation(cmannot);
516 * general purpose structure importer - designed to yield alignment useful for
517 * transfer of annotation to associated sequences
519 * @param alphaFoldCif
528 public static AlignmentI importDownloadedStructureFromUrl(
529 String alphaFoldCif, File tmpFile, String id, String chain,
530 String dbSource, String dbVersion) throws Exception
532 String file = tmpFile.getAbsolutePath();
533 // todo get rid of Type and use FileFormatI instead?
534 FileFormatI fileFormat = FileFormat.MMCif;
535 TFType tempfacType = TFType.PLDDT;
536 AlignmentI pdbAlignment = new FormatAdapter().readFile(tmpFile, file,
537 DataSourceType.FILE, fileFormat, tempfacType);
539 if (pdbAlignment != null)
541 List<SequenceI> toremove = new ArrayList<SequenceI>();
542 for (SequenceI pdbcs : pdbAlignment.getSequences())
546 for (PDBEntry pid : pdbcs.getAllPDBEntries())
548 if (pid.getFile() == file)
550 chid = pid.getChainCode();
553 if (chain == null || (chid != null && (chid.equals(chain)
554 || chid.trim().equals(chain.trim())
555 || (chain.trim().length() == 0 && chid.equals("_")))))
557 // FIXME seems to result in 'PDB|1QIP|1qip|A' - 1QIP is redundant.
558 // TODO: suggest simplify naming to 1qip|A as default name defined
559 pdbcs.setName(id + SEPARATOR + pdbcs.getName());
560 // Might need to add more metadata to the PDBEntry object
563 * PDBEntry entry = new PDBEntry(); // Construct the PDBEntry
564 * entry.setId(id); if (entry.getProperty() == null)
565 * entry.setProperty(new Hashtable());
566 * entry.getProperty().put("chains", pdbchain.id + "=" +
567 * sq.getStart() + "-" + sq.getEnd());
568 * sq.getDatasetSequence().addPDBId(entry);
571 // We make a DBRefEtntry because we have obtained the PDB file from
574 // JBPNote - PDB DBRefEntry should also carry the chain and mapping
576 if (dbSource != null)
578 DBRefEntry dbentry = new DBRefEntry(dbSource,
580 dbVersion, (chid == null ? id : id + chid));
582 pdbcs.addDBRef(dbentry);
583 // update any feature groups
584 List<SequenceFeature> allsf = pdbcs.getFeatures()
586 List<SequenceFeature> newsf = new ArrayList<SequenceFeature>();
587 if (allsf != null && allsf.size() > 0)
589 for (SequenceFeature f : allsf)
591 if (file.equals(f.getFeatureGroup()))
593 f = new SequenceFeature(f, f.type, f.begin, f.end, id,
598 pdbcs.setSequenceFeatures(newsf);
604 // mark this sequence to be removed from the alignment
605 // - since it's not from the right chain
609 // now remove marked sequences
610 for (SequenceI pdbcs : toremove)
612 pdbAlignment.deleteSequence(pdbcs);
613 if (pdbcs.getAnnotation() != null)
615 for (AlignmentAnnotation aa : pdbcs.getAnnotation())
617 pdbAlignment.deleteAnnotation(aa);
628 * @see jalview.ws.DbSourceProxy#isValidReference(java.lang.String)
631 public boolean isValidReference(String accession)
633 Regex r = getAccessionValidator();
634 return r.search(accession.trim());
641 public String getTestQuery()
643 return "AF-O15552-F1";
647 public String getDbName()
649 return "ALPHAFOLD"; // getDbSource();
659 * Returns a descriptor for suitable feature display settings with
661 * <li>ResNums or insertions features visible</li>
662 * <li>insertions features coloured red</li>
663 * <li>ResNum features coloured by label</li>
664 * <li>Insertions displayed above (on top of) ResNums</li>
668 public FeatureSettingsModelI getFeatureColourScheme()
670 return new PDBFeatureSettings();
674 private static final long PAE_CACHE_STALE_TIME = 1 * 86400000;
676 private static Map<String, File> paeDownloadCache = new HashMap<>();