3 import jalview.bin.Cache;
4 import jalview.datamodel.Alignment;
5 import jalview.datamodel.AlignmentAnnotation;
6 import jalview.datamodel.AlignmentI;
7 import jalview.datamodel.HiddenMarkovModel;
8 import jalview.datamodel.SequenceI;
9 import jalview.gui.AlignFrame;
10 import jalview.gui.Desktop;
11 import jalview.gui.JvOptionPane;
12 import jalview.io.DataSourceType;
13 import jalview.io.FileParse;
14 import jalview.io.StockholmFile;
15 import jalview.util.FileUtils;
16 import jalview.util.MessageManager;
17 import jalview.ws.params.ArgumentI;
18 import jalview.ws.params.simple.BooleanOption;
19 import jalview.ws.params.simple.Option;
21 import java.io.BufferedReader;
23 import java.io.FileReader;
24 import java.io.IOException;
25 import java.util.ArrayList;
26 import java.util.Collections;
27 import java.util.List;
28 import java.util.Scanner;
30 import javax.swing.JOptionPane;
32 public class HMMSearch extends HmmerCommand
34 static final String HMMSEARCH = "hmmsearch";
37 * constants for i18n lookup of passed parameter names
39 static final String DATABASE_KEY = "label.database";
41 static final String THIS_ALIGNMENT_KEY = "label.this_alignment";
43 static final String USE_ACCESSIONS_KEY = "label.use_accessions";
45 static final String AUTO_ALIGN_SEQS_KEY = "label.auto_align_seqs";
47 static final String NUMBER_OF_RESULTS_KEY = "label.number_of_results";
49 static final String TRIM_TERMINI_KEY = "label.trim_termini";
51 static final String REPORTING_CUTOFF_KEY = "label.reporting_cutoff";
53 static final String CUTOFF_NONE = "None";
55 static final String CUTOFF_SCORE = "Score";
57 static final String CUTOFF_EVALUE = "E-Value";
59 static final String SEQ_EVALUE_KEY = "label.seq_evalue";
61 static final String DOM_EVALUE_KEY = "label.dom_evalue";
63 static final String SEQ_SCORE_KEY = "label.seq_score";
65 static final String DOM_SCORE_KEY = "label.dom_score";
67 boolean realign = false;
71 int seqsToReturn = Integer.MAX_VALUE;
75 private String databaseName;
78 * Constructor for the HMMSearchThread
82 public HMMSearch(AlignFrame af, List<ArgumentI> args)
88 * Runs the HMMSearchThread: the data on the alignment or group is exported,
89 * then the command is executed in the command line and then the data is
90 * imported and displayed in a new frame. Call this method directly to execute
91 * synchronously, or via start() in a new Thread for asynchronously.
96 HiddenMarkovModel hmm = getHmmProfile();
99 // shouldn't happen if we got this far
100 Cache.log.error("Error: no hmm for hmmsearch");
104 SequenceI hmmSeq = hmm.getConsensusSequence();
105 long msgId = System.currentTimeMillis();
106 af.setProgressBar(MessageManager.getString("status.running_hmmsearch"),
111 File hmmFile = FileUtils.createTempFile("hmm", ".hmm");
112 File hitsAlignmentFile = FileUtils.createTempFile("hitAlignment",
114 File searchOutputFile = FileUtils.createTempFile("searchOutput",
117 exportHmm(hmm, hmmFile.getAbsoluteFile());
119 boolean ran = runCommand(searchOutputFile, hitsAlignmentFile, hmmFile);
122 JvOptionPane.showInternalMessageDialog(af, MessageManager
123 .formatMessage("warn.command_failed", "hmmsearch"));
127 importData(hmmSeq, hitsAlignmentFile, hmmFile, searchOutputFile);
128 // TODO make realignment of search results a step at this level
129 // and make it conditional on this.realign
130 } catch (IOException | InterruptedException e)
136 af.setProgressBar("", msgId);
141 * Executes an hmmsearch with the given hmm as input. The database to be
142 * searched is a local file as specified by the 'Database' parameter, or the
143 * current alignment (written to file) if none is specified.
145 * @param searchOutputFile
146 * @param hitsAlignmentFile
150 * @throws IOException
152 private boolean runCommand(File searchOutputFile, File hitsAlignmentFile,
153 File hmmFile) throws IOException
155 String command = getCommandPath(HMMSEARCH);
161 List<String> args = new ArrayList<>();
163 buildArguments(args, searchOutputFile, hitsAlignmentFile, hmmFile);
165 return runCommand(args);
169 * Appends command line arguments to the given list, to specify input and
170 * output files for the search, and any additional options that may have been
171 * passed from the parameters dialog
174 * @param searchOutputFile
175 * @param hitsAlignmentFile
177 * @throws IOException
179 protected void buildArguments(List<String> args, File searchOutputFile,
180 File hitsAlignmentFile, File hmmFile) throws IOException
183 args.add(getFilePath(searchOutputFile));
185 args.add(getFilePath(hitsAlignmentFile));
187 boolean dbFound = false;
189 File databaseFile = null;
191 boolean useEvalueCutoff = false;
192 boolean useScoreCutoff = false;
193 String seqEvalueCutoff = null;
194 String domEvalueCutoff = null;
195 String seqScoreCutoff = null;
196 String domScoreCutoff = null;
197 databaseName = "Alignment";
198 boolean searchAlignment = false;
202 for (ArgumentI arg : params)
204 String name = arg.getName();
205 if (MessageManager.getString(NUMBER_OF_RESULTS_KEY)
208 seqsToReturn = Integer.parseInt(arg.getValue());
210 else if (MessageManager.getString("action.search").equals(name))
212 searchAlignment = arg.getValue().equals(
213 MessageManager.getString(HMMSearch.THIS_ALIGNMENT_KEY));
215 else if (MessageManager.getString(DATABASE_KEY).equals(name))
217 dbPath = arg.getValue();
218 int pos = dbPath.lastIndexOf(File.separator);
219 databaseName = dbPath.substring(pos + 1);
220 databaseFile = new File(dbPath);
222 else if (MessageManager.getString(AUTO_ALIGN_SEQS_KEY)
227 else if (MessageManager.getString(USE_ACCESSIONS_KEY)
232 else if (MessageManager.getString(REPORTING_CUTOFF_KEY)
235 if (CUTOFF_EVALUE.equals(arg.getValue()))
237 useEvalueCutoff = true;
239 else if (CUTOFF_SCORE.equals(arg.getValue()))
241 useScoreCutoff = true;
244 else if (MessageManager.getString(SEQ_EVALUE_KEY).equals(name))
246 seqEvalueCutoff = arg.getValue();
248 else if (MessageManager.getString(SEQ_SCORE_KEY).equals(name))
250 seqScoreCutoff = arg.getValue();
252 else if (MessageManager.getString(DOM_EVALUE_KEY)
255 domEvalueCutoff = arg.getValue();
257 else if (MessageManager.getString(DOM_SCORE_KEY).equals(name))
259 domScoreCutoff = arg.getValue();
261 else if (MessageManager.getString(TRIM_TERMINI_KEY)
266 else if (MessageManager.getString(DATABASE_KEY).equals(name))
269 dbPath = arg.getValue();
270 if (!MessageManager.getString(THIS_ALIGNMENT_KEY)
273 int pos = dbPath.lastIndexOf(File.separator);
274 databaseName = dbPath.substring(pos + 1);
275 databaseFile = new File(dbPath);
284 args.add(seqEvalueCutoff);
286 args.add(domEvalueCutoff);
288 else if (useScoreCutoff)
291 args.add(seqScoreCutoff);
293 args.add(domScoreCutoff);
296 // if (!dbFound || MessageManager.getString(THIS_ALIGNMENT_KEY)
301 * no external database specified for search, so
302 * export current alignment as 'database' to search,
303 * excluding any HMM consensus sequences it contains
305 databaseFile = FileUtils.createTempFile("database", ".sto");
306 AlignmentI al = af.getViewport().getAlignment();
307 AlignmentI copy = new Alignment(al);
308 List<SequenceI> hmms = copy.getHmmSequences();
309 for (SequenceI hmmSeq : hmms)
311 copy.deleteSequence(hmmSeq);
313 exportStockholm(copy.getSequencesArray(), databaseFile, null);
316 args.add(getFilePath(hmmFile));
317 args.add(getFilePath(databaseFile));
321 * Imports the data from the temporary file to which the output of hmmsearch
322 * was directed. The results are optionally realigned using hmmalign.
326 private void importData(SequenceI hmmSeq, File inputAlignmentTemp,
327 File hmmTemp, File searchOutputFile)
328 throws IOException, InterruptedException
330 BufferedReader br = new BufferedReader(
331 new FileReader(inputAlignmentTemp));
334 if (br.readLine() == null)
336 JOptionPane.showMessageDialog(af,
337 MessageManager.getString("label.no_sequences_found"));
340 StockholmFile file = new StockholmFile(new FileParse(
341 inputAlignmentTemp.getAbsolutePath(), DataSourceType.FILE));
342 seqs = file.getSeqsAsArray();
344 readTable(searchOutputFile);
346 int seqCount = Math.min(seqs.length, seqsToReturn);
347 SequenceI[] hmmAndSeqs = new SequenceI[seqCount + 1];
348 hmmAndSeqs[0] = hmmSeq;
349 System.arraycopy(seqs, 0, hmmAndSeqs, 1, seqCount);
353 realignResults(hmmAndSeqs);
357 AlignmentI al = new Alignment(hmmAndSeqs);
358 AlignFrame alignFrame = new AlignFrame(al, AlignFrame.DEFAULT_WIDTH,
359 AlignFrame.DEFAULT_HEIGHT);
360 String ttl = "hmmSearch of " + databaseName + " using "
362 Desktop.addInternalFrame(alignFrame, ttl, AlignFrame.DEFAULT_WIDTH,
363 AlignFrame.DEFAULT_HEIGHT);
367 inputAlignmentTemp.delete();
368 searchOutputFile.delete();
379 * Realigns the given sequences using hmmalign, to the HMM profile sequence
380 * which is the first in the array, and opens the results in a new frame
384 protected void realignResults(SequenceI[] hmmAndSeqs)
387 * and align the search results to the HMM profile
389 AlignmentI al = new Alignment(hmmAndSeqs);
390 AlignFrame frame = new AlignFrame(al, 1, 1);
391 List<ArgumentI> alignArgs = new ArrayList<>();
392 String alignTo = hmmAndSeqs[0].getName();
393 List<String> options = Collections.singletonList(alignTo);
394 Option option = new Option(MessageManager.getString("label.use_hmm"),
395 "", true, alignTo, alignTo, options, null);
396 alignArgs.add(option);
399 alignArgs.add(new BooleanOption(
400 MessageManager.getString(TRIM_TERMINI_KEY),
401 MessageManager.getString("label.trim_termini_desc"), true,
404 HmmerCommand hmmalign = new HMMAlign(frame, alignArgs);
409 * Reads in the scores table output by hmmsearch and adds annotation to
410 * sequences for E-value and bit score
412 * @param inputTableTemp
413 * @throws IOException
415 void readTable(File inputTableTemp) throws IOException
417 BufferedReader br = new BufferedReader(new FileReader(inputTableTemp));
419 while (!line.startsWith("Query:"))
421 line = br.readLine();
423 for (int i = 0; i < 5; i++)
425 line = br.readLine();
429 while (!" ------ inclusion threshold ------".equals(line)
432 SequenceI seq = seqs[index];
433 Scanner scanner = new Scanner(line);
434 String str = scanner.next();
435 addScoreAnnotation(str, seq, "hmmsearch E-value",
436 "Full sequence E-value");
437 str = scanner.next();
438 addScoreAnnotation(str, seq, "hmmsearch Score",
439 "Full sequence bit score");
441 line = br.readLine();
449 * A helper method that adds one score-only (non-positional) annotation to a
457 protected void addScoreAnnotation(String value, SequenceI seq,
458 String label, String description)
462 AlignmentAnnotation annot = new AlignmentAnnotation(label,
464 annot.setCalcId(HMMSEARCH);
465 double eValue = Double.parseDouble(value);
466 annot.setScore(eValue);
467 annot.setSequenceRef(seq);
468 seq.addAlignmentAnnotation(annot);
469 } catch (NumberFormatException e)
471 System.err.println("Error parsing " + label + " from " + value);