3 import jalview.datamodel.HMMNode;
4 import jalview.datamodel.HiddenMarkovModel;
5 import jalview.datamodel.SequenceI;
7 import java.io.BufferedReader;
8 import java.io.FileNotFoundException;
9 import java.io.IOException;
10 import java.io.PrintWriter;
11 import java.io.UnsupportedEncodingException;
12 import java.util.ArrayList;
13 import java.util.List;
14 import java.util.Scanner;
18 * Adds capability to read in and write out HMMER3 files. Currently only supports HMMER3/f.
24 public class HMMFile extends AlignFile
25 implements AlignmentFileReaderI, AlignmentFileWriterI
27 // HMM to store file data
28 private HiddenMarkovModel hmm = new HiddenMarkovModel();
33 // number of possible transitions
34 private final int NUMBER_OF_TRANSITIONS = 7;
36 private final String NEW_LINE = "\n";
41 //number of symbols in the alphabet used in the hidden Markov model
44 private final String SPACE = " ";
46 private final String COMPO = "COMPO";
48 private final String EMPTY = "";
50 //This is a line that needs to be added to each HMMER£ file. It is purely for readability.
51 private static final String TRANSITIONTYPELINE = "m->m m->i m->d i->m i->i d->m d->d";
54 * Constructor for HMMFile
58 public HMMFile(FileParse source) throws IOException
64 * Default constructor, do not use!
72 * Returns the HMM produced by reading in a HMMER3 file.
76 public HiddenMarkovModel getHMM()
82 * Sets the HMM used in this file.
86 public void setHMM(HiddenMarkovModel model)
92 * Gets the name of the hidden Markov model.
96 public String getName()
102 * Reads the data from HMM file into the HMM field on this object.
104 * @throws IOException
107 public void parse() throws IOException
109 parseFileProperties(dataIn);
114 * Reads the data from HMM file into the HMM field on this object.
116 * @throws IOException
119 public void parse(BufferedReader br) throws IOException
121 parseFileProperties(br);
128 * Imports the file properties from a HMMER3 file.
131 * The buffered reader used to read in the file.
132 * @throws IOException
134 void parseFileProperties(BufferedReader input) throws IOException
136 boolean readingFile = true;
137 fileHeader = input.readLine();
138 String line = input.readLine();
143 Scanner parser = new Scanner(line);
144 String next = parser.next();
145 if ("HMM".equals(next)) // indicates start of HMM data (end of file
149 hmm.fillSymbols(parser);
150 numberOfSymbols = hmm.getNumberOfSymbols();
152 else if ("STATS".equals(next))
158 value = parser.next() + SPACE + SPACE + parser.next();
159 hmm.addFileProperty(key, value);
164 String value = parser.next();
165 while (parser.hasNext())
167 value = value + SPACE + parser.next();
169 hmm.addFileProperty(key, value);
173 line = input.readLine();
183 * Parses the model data from the HMMER3 file
186 * The buffered reader used to read the file.
187 * @throws IOException
189 void parseModel(BufferedReader input) throws IOException
191 String line = input.readLine();
193 while (!"//".equals(line))
195 hmm.getNodes().add(new HMMNode());
197 Scanner matchReader = new Scanner(line);
198 next = matchReader.next();
199 if (next.equals(COMPO) || node > 0)
201 // stores match emission line in list
202 List<Double> matches = new ArrayList<>();
203 matches = fillList(matchReader, numberOfSymbols);
204 hmm.getNodes().get(node).setMatchEmissions(matches);
207 parseAnnotations(matchReader, node);
211 // stores insert emission line in list
212 line = input.readLine();
213 Scanner insertReader = new Scanner(line);
214 List<Double> inserts = new ArrayList<>();
215 inserts = fillList(insertReader, numberOfSymbols);
216 hmm.getNodes().get(node).setInsertEmissions(inserts);
217 insertReader.close();
219 // stores state transition line in list
220 line = input.readLine();
221 Scanner transitionReader = new Scanner(line);
222 List<Double> transitions = new ArrayList<>();
223 transitions = fillList(transitionReader, NUMBER_OF_TRANSITIONS);
224 hmm.getNodes().get(node).setStateTransitions(transitions);
225 transitionReader.close();
226 line = input.readLine();
233 * Parses the annotations on the match emission line.
236 * The scanner which is processing match emission line.
238 * The index of node which is being scanned.
240 void parseAnnotations(Scanner scanner, int index)
242 if (hmm.mapIsActive())
245 column = scanner.nextInt();
246 hmm.getNodes().get(index).setAlignmentColumn(column);
247 hmm.getNodeLookup().put(column, index);
254 if (scanner.hasNext())
257 consensusR = charValue(scanner.next());
258 hmm.getNodes().get(index).setConsensusResidue(consensusR);
261 if (scanner.hasNext())
264 reference = charValue(scanner.next());
265 hmm.getNodes().get(index).setReferenceAnnotation(reference);
268 if (scanner.hasNext())
271 value = charValue(scanner.next());
272 hmm.getNodes().get(index).setMaskValue(value);
274 if (scanner.hasNext())
277 consensusS = charValue(scanner.next());
278 hmm.getNodes().get(index).setConsensusStructure(consensusS);
285 * Fills a list of doubles based on an input line.
288 * The scanner for the line containing the data to be transferred to
290 * @param numberOfElements
291 * The number of elements in the list to be filled.
292 * @return filled list Returns the list of doubles.
294 static List<Double> fillList(Scanner input,
295 int numberOfElements)
297 List<Double> list = new ArrayList<>();
298 for (int i = 0; i < numberOfElements; i++)
301 String next = input.next();
302 if (next.contains("*")) // state transitions to or from delete states
303 // occasionally have values of -infinity. These
304 // values are represented by an * in the .hmm
307 list.add(Double.NEGATIVE_INFINITY);
311 double prob = Double.valueOf(next);
312 prob = Math.pow(Math.E, -prob);
321 * Writes a HMM to a file/
323 * @param exportLocation
324 * Filename, URL or Pasted String to write to.
325 * @throws FileNotFoundException
326 * @throws UnsupportedEncodingException
330 public void exportFile(String exportLocation) throws IOException
332 StringBuilder file = new StringBuilder();
333 appendFileProperties(file);
337 PrintWriter output = new PrintWriter(exportLocation);
344 * Returns a string to be added to the StringBuilder containing the entire
347 * @param initialColumnSeparation
348 * The initial whitespace separation between the left side of the
349 * file and first character.
350 * @param columnSeparation
351 * The separation between subsequent data entries.
353 * The list fo data to be added to the String.
356 String addData(int initialColumnSeparation,
357 int columnSeparation, List<String> data)
361 for (String value : data)
365 line += String.format("%" + initialColumnSeparation + "s", value);
369 line += String.format("%" + columnSeparation + "s", value);
377 * Converts list of characters into a list of Strings.
380 * @return Returns the list of Strings.
382 List<String> charListToStringList(List<Character> list)
384 List<String> strList = new ArrayList<>();
385 for (char value : list)
387 String strValue = Character.toString(value);
388 strList.add(strValue);
394 * Converts a list of doubles into a list of Strings, rounded to the nearest
398 * @param noOfDecimals
401 List<String> doubleListToStringList(List<Double> list)
403 List<String> strList = new ArrayList<>();
404 for (double value : list)
409 strValue = String.format("%.5f", value);
412 else if (value == -0.00000d)
414 strValue = "0.00000";
421 strList.add(strValue);
427 * Converts a primitive array of Strings to a list of Strings.
432 List<String> stringArrayToStringList(String[] array)
434 List<String> list = new ArrayList<>();
435 for (String value : array)
444 * Appends the hidden Markov model data to the StringBuilder containing the
448 * The StringBuilder containing the output.
450 void appendModel(StringBuilder file)
452 String symbolLine = "HMM";
453 List<Character> charSymbols = hmm.getSymbols();
454 List<String> strSymbols;
455 strSymbols = charListToStringList(charSymbols);
456 symbolLine += addData(11, 9, strSymbols);
457 file.append(symbolLine + NEW_LINE);
458 file.append(TRANSITIONTYPELINE + NEW_LINE);
460 int length = hmm.getLength();
462 for (int node = 0; node <= length; node++)
467 matchLine = String.format("%7s", "COMPO");
471 matchLine = String.format("%7s", node);
474 List<String> strMatches;
475 List<Double> doubleMatches;
476 doubleMatches = hmm.getNode(node).getMatchEmissions();
477 convertListToLogSpace(doubleMatches);
478 strMatches = doubleListToStringList(doubleMatches);
479 matchLine += addData(10, 9, strMatches);
484 matchLine += SPACE + hmm.getNodeAlignmentColumn(node);
485 matchLine += SPACE + hmm.getConsensusResidue(node);
486 matchLine += SPACE + hmm.getReferenceAnnotation(node);
487 matchLine += SPACE + hmm.getMaskedValue(node);
488 matchLine += SPACE + hmm.getConsensusStructure(node);
492 file.append(matchLine + NEW_LINE);
494 String insertLine = EMPTY;
495 List<String> strInserts;
496 List<Double> doubleInserts;
497 doubleInserts = hmm.getNode(node).getInsertEmissions();
498 convertListToLogSpace(doubleInserts);
499 strInserts = doubleListToStringList(doubleInserts);
500 insertLine += addData(17, 9, strInserts);
502 file.append(insertLine + NEW_LINE);
504 String transitionLine = EMPTY;
505 List<String> strTransitions;
506 List<Double> doubleTransitions;
507 doubleTransitions = hmm.getNode(node).getStateTransitions();
508 convertListToLogSpace(doubleTransitions);
509 strTransitions = doubleListToStringList(doubleTransitions);
510 transitionLine += addData(17, 9, strTransitions);
512 file.append(transitionLine + NEW_LINE);
517 * Appends the hidden Markov model file properties to the StringBuilder
518 * containing the output
521 * The StringBuilder containing the output.
523 void appendFileProperties(StringBuilder file)
527 file.append(fileHeader + NEW_LINE);
529 line = String.format("%-5s %1s", "NAME", hmm.getName());
530 file.append((line + NEW_LINE));
532 if (hmm.getAccessionNumber() != null)
534 line = String.format("%-5s %1s", "ACC", hmm.getAccessionNumber());
535 file.append((line + NEW_LINE));
538 if (hmm.getDescription() != null)
540 line = String.format("%-5s %1s", "DESC", hmm.getDescription());
541 file.append((line + NEW_LINE));
543 line = String.format("%-5s %1s", "LENG", hmm.getLength());
544 file.append((line + NEW_LINE));
546 if (hmm.getMaxInstanceLength() != null)
548 line = String.format("%-5s %1s", "MAXL", hmm.getMaxInstanceLength());
549 file.append((line + NEW_LINE));
551 line = String.format("%-5s %1s", "ALPH", hmm.getAlphabetType());
552 file.append((line + NEW_LINE));
557 status = hmm.referenceAnnotationIsActive();
558 statusStr = HiddenMarkovModel.findStringFromBoolean(status);
559 line = String.format("%-5s %1s", "RF",
561 file.append((line + NEW_LINE));
563 status = hmm.maskValueIsActive();
564 statusStr = HiddenMarkovModel.findStringFromBoolean(status);
565 line = String.format("%-5s %1s", "MM",
567 file.append((line + NEW_LINE));
569 status = hmm.consensusResidueIsActive();
570 statusStr = HiddenMarkovModel.findStringFromBoolean(status);
571 line = String.format("%-5s %1s", "CONS",
573 file.append((line + NEW_LINE));
575 status = hmm.consensusStructureIsActive();
576 statusStr = HiddenMarkovModel.findStringFromBoolean(status);
577 line = String.format("%-5s %1s", "CS",
579 file.append((line + NEW_LINE));
581 status = hmm.mapIsActive();
582 statusStr = HiddenMarkovModel.findStringFromBoolean(status);
583 line = String.format("%-5s %1s", "MAP",
585 file.append((line + NEW_LINE));
588 if (hmm.getDate() != null)
590 line = String.format("%-5s %1s", "DATE", hmm.getDate());
591 file.append((line + NEW_LINE));
593 if (hmm.getNumberOfSequences() != null)
595 line = String.format("%-5s %1s", "NSEQ", hmm.getNumberOfSequences());
596 file.append((line + NEW_LINE));
598 if (hmm.getEffectiveNumberOfSequences() != null)
600 line = String.format("%-5s %1s", "EFFN",
601 hmm.getEffectiveNumberOfSequences());
602 file.append((line + NEW_LINE));
604 if (hmm.getCheckSum() != null)
606 line = String.format("%-5s %1s", "CKSUM", hmm.getCheckSum());
607 file.append((line + NEW_LINE));
609 if (hmm.getGatheringThreshold() != null)
611 line = String.format("%-5s %1s", "GA", hmm.getGatheringThreshold());
612 file.append((line + NEW_LINE));
615 if (hmm.getTrustedCutoff() != null)
617 line = String.format("%-5s %1s", "TC", hmm.getTrustedCutoff());
618 file.append((line + NEW_LINE));
620 if (hmm.getNoiseCutoff() != null)
622 line = String.format("%-5s %1s", "NC", hmm.getNoiseCutoff());
623 file.append((line + NEW_LINE));
625 if (hmm.getMSV() != null)
627 line = String.format("%-19s %18s", "STATS LOCAL MSV", hmm.getMSV());
628 file.append((line + NEW_LINE));
630 line = String.format("%-19s %18s", "STATS LOCAL VITERBI",
632 file.append((line + NEW_LINE));
634 line = String.format("%-19s %18s", "STATS LOCAL FORWARD",
636 file.append((line + NEW_LINE));
642 * Returns the char value of a single lettered String.
647 char charValue(String string)
650 character = string.charAt(0);
656 public String print(SequenceI[] seqs, boolean jvsuffix)
663 * Converts the probabilities contained in a list into log space.
667 void convertListToLogSpace(List<Double> list)
670 for (int i = 0; i < list.size(); i++)
672 double prob = list.get(i);
673 double logProb = -1 * Math.log(prob);
675 list.set(i, logProb);