datamodel/compbio/data/sequence/SequenceUtil.java

   1 /*\r
   2  * @(#)SequenceUtil.java 1.0 September 2009 Copyright (c) 2009 Peter Troshin\r
   3  * Jalview Web Services version: 2.0 This library is free software; you can\r
   4  * redistribute it and/or modify it under the terms of the Apache License\r
   5  * version 2 as published by the Apache Software Foundation This library is\r
   6  * distributed in the hope that it will be useful, but WITHOUT ANY WARRANTY;\r
   7  * without even the implied warranty of MERCHANTABILITY or FITNESS FOR A\r
   8  * PARTICULAR PURPOSE. See the Apache License for more details. A copy of the\r
   9  * license is in apache_license.txt. It is also available here: see:\r
  10  * http://www.apache.org/licenses/LICENSE-2.0.txt Any republication or derived\r
  11  * work distributed in source code form must include this copyright and license\r
  12  * notice.\r
  13  */\r
  14 \r
  15 package compbio.data.sequence;\r
  16 \r
  17 import java.io.BufferedReader;\r
  18 import java.io.BufferedWriter;\r
  19 import java.io.Closeable;\r
  20 import java.io.File;\r
  21 import java.io.FileInputStream;\r
  22 import java.io.IOException;\r
  23 import java.io.InputStream;\r
  24 import java.io.InputStreamReader;\r
  25 import java.io.OutputStream;\r
  26 import java.io.OutputStreamWriter;\r
  27 import java.util.ArrayList;\r
  28 import java.util.HashSet;\r
  29 import java.util.List;\r
  30 import java.util.Scanner;\r
  31 import java.util.logging.Level;\r
  32 import java.util.regex.Matcher;\r
  33 import java.util.regex.Pattern;\r
  34 \r
  35 /**\r
  36  * Utility class for operations on sequences\r
  37  * \r
  38  * @author Petr Troshin\r
  39  * @version 1.0\r
  40  */\r
  41 public final class SequenceUtil {\r
  42 \r
  43         /**\r
  44          * A whitespace character: [\t\n\x0B\f\r]\r
  45          */\r
  46         public static final Pattern WHITE_SPACE = Pattern.compile("\\s");\r
  47 \r
  48         /**\r
  49          * A digit\r
  50          */\r
  51         public static final Pattern DIGIT = Pattern.compile("\\d");\r
  52 \r
  53         /**\r
  54          * Non word\r
  55          */\r
  56         public static final Pattern NONWORD = Pattern.compile("\\W");\r
  57 \r
  58         /**\r
  59          * Valid Amino acids\r
  60          */\r
  61         public static final Pattern AA = Pattern.compile("[ARNDCQEGHILKMFPSTWYV]+",\r
  62                         Pattern.CASE_INSENSITIVE);\r
  63 \r
  64         /**\r
  65          * inversion of AA pattern\r
  66          */\r
  67         public static final Pattern NON_AA = Pattern.compile(\r
  68                         "[^ARNDCQEGHILKMFPSTWYV]+", Pattern.CASE_INSENSITIVE);\r
  69 \r
  70         /**\r
  71          * Same as AA pattern but with two additional letters - XU\r
  72          */\r
  73         public static final Pattern AMBIGUOUS_AA = Pattern.compile(\r
  74                         "[ARNDCQEGHILKMFPSTWYVXU]+", Pattern.CASE_INSENSITIVE);\r
  75 \r
  76         /**\r
  77          * Nucleotides a, t, g, c, u\r
  78          */\r
  79         public static final Pattern NUCLEOTIDE = Pattern.compile("[AGTCU]+",\r
  80                         Pattern.CASE_INSENSITIVE);\r
  81 \r
  82         /**\r
  83          * Ambiguous nucleotide\r
  84          */\r
  85         public static final Pattern AMBIGUOUS_NUCLEOTIDE = Pattern.compile(\r
  86                         "[AGTCRYMKSWHBVDNU]+", Pattern.CASE_INSENSITIVE); // see IUPAC\r
  87         /**\r
  88          * Non nucleotide\r
  89          */\r
  90         public static final Pattern NON_NUCLEOTIDE = Pattern.compile("[^AGTCU]+",\r
  91                         Pattern.CASE_INSENSITIVE);\r
  92 \r
  93         private SequenceUtil() {\r
  94         } // utility class, no instantiation\r
  95 \r
  96         /*\r
  97          * public static void write_PirSeq(OutputStream os, FastaSequence seq)\r
  98          * throws IOException { BufferedWriter pir_out = new BufferedWriter(new\r
  99          * OutputStreamWriter(os)); pir_out.write(">P1;" + seq.getId() +\r
 100          * SysPrefs.newlinechar); pir_out.write(seq.getSequence() +\r
 101          * SysPrefs.newlinechar); pir_out.close(); } public static void\r
 102          * write_FastaSeq(OutputStream os, FastaSequence seq) throws IOException {\r
 103          * BufferedWriter fasta_out = new BufferedWriter( new\r
 104          * OutputStreamWriter(os)); fasta_out.write(">" + seq.getId() +\r
 105          * SysPrefs.newlinechar); fasta_out.write(seq.getSequence() +\r
 106          * SysPrefs.newlinechar); fasta_out.close(); }\r
 107          */\r
 108 \r
 109         /**\r
 110          * @return true is the sequence contains only letters a,c, t, g, u\r
 111          */\r
 112         public static boolean isNucleotideSequence(final FastaSequence s) {\r
 113                 return SequenceUtil.isNonAmbNucleotideSequence(s.getSequence());\r
 114         }\r
 115 \r
 116         /**\r
 117          * Ambiguous DNA chars : AGTCRYMKSWHBVDN // differs from protein in only one\r
 118          * (!) - B char\r
 119          */\r
 120         public static boolean isNonAmbNucleotideSequence(String sequence) {\r
 121                 sequence = SequenceUtil.cleanSequence(sequence);\r
 122                 if (SequenceUtil.DIGIT.matcher(sequence).find()) {\r
 123                         return false;\r
 124                 }\r
 125                 if (SequenceUtil.NON_NUCLEOTIDE.matcher(sequence).find()) {\r
 126                         return false;\r
 127                         /*\r
 128                          * System.out.format("I found the text starting at " +\r
 129                          * "index %d and ending at index %d.%n", nonDNAmatcher .start(),\r
 130                          * nonDNAmatcher.end());\r
 131                          */\r
 132                 }\r
 133                 final Matcher DNAmatcher = SequenceUtil.NUCLEOTIDE.matcher(sequence);\r
 134                 return DNAmatcher.find();\r
 135         }\r
 136 \r
 137         /**\r
 138          * Removes all whitespace chars in the sequence string\r
 139          * \r
 140          * @param sequence\r
 141          * @return cleaned up sequence\r
 142          */\r
 143         public static String cleanSequence(String sequence) {\r
 144                 assert sequence != null;\r
 145                 final Matcher m = SequenceUtil.WHITE_SPACE.matcher(sequence);\r
 146                 sequence = m.replaceAll("").toUpperCase();\r
 147                 return sequence;\r
 148         }\r
 149 \r
 150         /**\r
 151          * Removes all special characters and digits as well as whitespace chars\r
 152          * from the sequence\r
 153          * \r
 154          * @param sequence\r
 155          * @return cleaned up sequence\r
 156          */\r
 157         public static String deepCleanSequence(String sequence) {\r
 158                 sequence = SequenceUtil.cleanSequence(sequence);\r
 159                 sequence = SequenceUtil.DIGIT.matcher(sequence).replaceAll("");\r
 160                 sequence = SequenceUtil.NONWORD.matcher(sequence).replaceAll("");\r
 161                 final Pattern othernonSeqChars = Pattern.compile("[_-]+");\r
 162                 sequence = othernonSeqChars.matcher(sequence).replaceAll("");\r
 163                 return sequence;\r
 164         }\r
 165 \r
 166         /**\r
 167          * @param sequence\r
 168          * @return true is the sequence is a protein sequence, false overwise\r
 169          */\r
 170         public static boolean isProteinSequence(String sequence) {\r
 171                 sequence = SequenceUtil.cleanSequence(sequence);\r
 172                 if (SequenceUtil.isNonAmbNucleotideSequence(sequence)) {\r
 173                         return false;\r
 174                 }\r
 175                 if (SequenceUtil.DIGIT.matcher(sequence).find()) {\r
 176                         return false;\r
 177                 }\r
 178                 if (SequenceUtil.NON_AA.matcher(sequence).find()) {\r
 179                         return false;\r
 180                 }\r
 181                 final Matcher protmatcher = SequenceUtil.AA.matcher(sequence);\r
 182                 return protmatcher.find();\r
 183         }\r
 184 \r
 185         /**\r
 186          * Check whether the sequence confirms to amboguous protein sequence\r
 187          * \r
 188          * @param sequence\r
 189          * @return return true only if the sequence if ambiguous protein sequence\r
 190          *         Return false otherwise. e.g. if the sequence is non-ambiguous\r
 191          *         protein or DNA\r
 192          */\r
 193         public static boolean isAmbiguosProtein(String sequence) {\r
 194                 sequence = SequenceUtil.cleanSequence(sequence);\r
 195                 if (SequenceUtil.isNonAmbNucleotideSequence(sequence)) {\r
 196                         return false;\r
 197                 }\r
 198                 if (SequenceUtil.DIGIT.matcher(sequence).find()) {\r
 199                         return false;\r
 200                 }\r
 201                 if (SequenceUtil.NON_AA.matcher(sequence).find()) {\r
 202                         return false;\r
 203                 }\r
 204                 if (SequenceUtil.AA.matcher(sequence).find()) {\r
 205                         return false;\r
 206                 }\r
 207                 final Matcher amb_prot = SequenceUtil.AMBIGUOUS_AA.matcher(sequence);\r
 208                 return amb_prot.find();\r
 209         }\r
 210 \r
 211         /**\r
 212          * Writes list of FastaSequeces into the outstream formatting the sequence\r
 213          * so that it contains width chars on each line\r
 214          * \r
 215          * @param outstream\r
 216          * @param sequences\r
 217          * @param width\r
 218          *            - the maximum number of characters to write in one line\r
 219          * @throws IOException\r
 220          */\r
 221         public static void writeFasta(final OutputStream outstream,\r
 222                         final List<FastaSequence> sequences, final int width)\r
 223                         throws IOException {\r
 224                 writeFastaKeepTheStream(outstream, sequences, width);\r
 225                 outstream.close();\r
 226         }\r
 227 \r
 228         public static void writeFastaKeepTheStream(final OutputStream outstream,\r
 229                         final List<FastaSequence> sequences, final int width)\r
 230                         throws IOException {\r
 231                 final OutputStreamWriter writer = new OutputStreamWriter(outstream);\r
 232                 final BufferedWriter fastawriter = new BufferedWriter(writer);\r
 233                 for (final FastaSequence fs : sequences) {\r
 234                         fastawriter.write(">" + fs.getId() + "\n");\r
 235                         fastawriter.write(fs.getFormatedSequence(width));\r
 236                         fastawriter.write("\n");\r
 237                 }\r
 238                 fastawriter.flush();\r
 239                 writer.flush();\r
 240         }\r
 241 \r
 242         /**\r
 243          * Reads fasta sequences from inStream into the list of FastaSequence\r
 244          * objects\r
 245          * \r
 246          * @param inStream\r
 247          *            from\r
 248          * @return list of FastaSequence objects\r
 249          * @throws IOException\r
 250          */\r
 251         public static List<FastaSequence> readFasta(final InputStream inStream)\r
 252                         throws IOException {\r
 253                 final List<FastaSequence> seqs = new ArrayList<FastaSequence>();\r
 254 \r
 255                 final BufferedReader infasta = new BufferedReader(\r
 256                                 new InputStreamReader(inStream, "UTF8"), 16000);\r
 257                 final Pattern pattern = Pattern.compile("//s+");\r
 258 \r
 259                 String line;\r
 260                 String sname = "", seqstr = null;\r
 261                 do {\r
 262                         line = infasta.readLine();\r
 263                         if ((line == null) || line.startsWith(">")) {\r
 264                                 if (seqstr != null) {\r
 265                                         seqs.add(new FastaSequence(sname.substring(1), seqstr));\r
 266                                 }\r
 267                                 sname = line; // remove >\r
 268                                 seqstr = "";\r
 269                         } else {\r
 270                                 final String subseq = pattern.matcher(line).replaceAll("");\r
 271                                 seqstr += subseq;\r
 272                         }\r
 273                 } while (line != null);\r
 274 \r
 275                 infasta.close();\r
 276                 return seqs;\r
 277         }\r
 278 \r
 279         /**\r
 280          * Writes FastaSequence in the file, each sequence will take one line only\r
 281          * \r
 282          * @param os\r
 283          * @param sequences\r
 284          * @throws IOException\r
 285          */\r
 286         public static void writeFasta(final OutputStream os,\r
 287                         final List<FastaSequence> sequences) throws IOException {\r
 288                 final OutputStreamWriter outWriter = new OutputStreamWriter(os);\r
 289                 final BufferedWriter fasta_out = new BufferedWriter(outWriter);\r
 290                 for (final FastaSequence fs : sequences) {\r
 291                         fasta_out.write(fs.getOnelineFasta());\r
 292                 }\r
 293                 fasta_out.close();\r
 294                 outWriter.close();\r
 295         }\r
 296 \r
 297         public static List<AnnotatedSequence> readJRonn(final File result)\r
 298                         throws IOException, UnknownFileFormatException {\r
 299                 InputStream input = new FileInputStream(result);\r
 300                 List<AnnotatedSequence> sequences = readJRonn(input);\r
 301                 input.close();\r
 302                 return sequences;\r
 303         }\r
 304 \r
 305         /**\r
 306          * Reader for JRonn horizontal file format\r
 307          * \r
 308          * <pre>\r
 309          * &gtFoobar M G D T T A G 0.48 0.42\r
 310          * 0.42 0.48 0.52 0.53 0.54\r
 311          * \r
 312          * <pre>\r
 313          * Where all values are tab delimited\r
 314          * \r
 315          * @param inStream\r
 316          *            the InputStream connected to the JRonn output file\r
 317          * @return List of {@link AnnotatedSequence} objects\r
 318          * @throws IOException\r
 319          *             is thrown if the inStream has problems accessing the data\r
 320          * @throws UnknownFileFormatException\r
 321          *             is thrown if the inStream represents an unknown source of\r
 322          * data, i.e. not a JRonn output\r
 323          */\r
 324         public static List<AnnotatedSequence> readJRonn(final InputStream inStream)\r
 325                         throws IOException, UnknownFileFormatException {\r
 326                 final List<AnnotatedSequence> seqs = new ArrayList<AnnotatedSequence>();\r
 327 \r
 328                 final BufferedReader infasta = new BufferedReader(\r
 329                                 new InputStreamReader(inStream, "UTF8"), 16000);\r
 330 \r
 331                 String line;\r
 332                 String sname = "";\r
 333                 do {\r
 334                         line = infasta.readLine();\r
 335                         if (line == null || line.isEmpty()) {\r
 336                                 // skip empty lines\r
 337                                 continue;\r
 338                         }\r
 339                         if (line.startsWith(">")) {\r
 340                                 // read name\r
 341                                 sname = line.trim().substring(1);\r
 342                                 // read sequence line\r
 343                                 line = infasta.readLine();\r
 344                                 final String sequence = line.replace("\t", "");\r
 345                                 // read annotation line\r
 346                                 line = infasta.readLine();\r
 347                                 String[] annotValues = line.split("\t");\r
 348                                 float[] annotation = convertToNumber(annotValues);\r
 349                                 if (annotation.length != sequence.length()) {\r
 350                                         throw new UnknownFileFormatException(\r
 351                                                         "File does not look like Jronn horizontally formatted output file!\n"\r
 352                                                                         + JRONN_WRONG_FORMAT_MESSAGE);\r
 353                                 }\r
 354                                 seqs.add(new AnnotatedSequence(sname, sequence, annotation));\r
 355                         }\r
 356                 } while (line != null);\r
 357 \r
 358                 infasta.close();\r
 359                 return seqs;\r
 360         }\r
 361 \r
 362         private static float[] convertToNumber(String[] annotValues)\r
 363                         throws UnknownFileFormatException {\r
 364                 float[] annotation = new float[annotValues.length];\r
 365                 try {\r
 366                         for (int i = 0; i < annotation.length; i++) {\r
 367                                 annotation[i] = Float.parseFloat(annotValues[i]);\r
 368                         }\r
 369                 } catch (NumberFormatException e) {\r
 370                         throw new UnknownFileFormatException(JRONN_WRONG_FORMAT_MESSAGE,\r
 371                                         e.getCause());\r
 372                 }\r
 373                 return annotation;\r
 374         }\r
 375 \r
 376         private static final String JRONN_WRONG_FORMAT_MESSAGE = "Jronn file must be in the following format:\n"\r
 377                         + ">sequence_name\n "\r
 378                         + "M    V       S\n"\r
 379                         + "0.43 0.22    0.65\n"\r
 380                         + "Where first line is the sequence name,\n"\r
 381                         + "second line is the tab delimited sequence,\n"\r
 382                         + "third line contains tab delimited disorder prediction values.\n"\r
 383                         + "No lines are allowed between these three. Additionally, the number of  "\r
 384                         + "sequence residues must be equal to the number of the disorder values.";\r
 385 \r
 386         /**\r
 387          * Closes the Closable and logs the exception if any\r
 388          * \r
 389          * @param log\r
 390          * @param stream\r
 391          */\r
 392         public final static void closeSilently(java.util.logging.Logger log,\r
 393                         Closeable stream) {\r
 394                 if (stream != null) {\r
 395                         try {\r
 396                                 stream.close();\r
 397                         } catch (IOException e) {\r
 398                                 log.log(Level.WARNING, e.getLocalizedMessage(), e.getCause());\r
 399                         }\r
 400                 }\r
 401         }\r
 402 \r
 403         /**\r
 404          * \r
 405          * TODO complete!\r
 406          * \r
 407          * # RESIDUE COILS REM465 HOTLOOPS M 0.86010 0.88512 0.37094 T 0.79983\r
 408          * 0.85864 0.44331 .... # RESIDUE COILS REM465 HOTLOOPS M 0.86010 0.88512\r
 409          * 0.37094\r
 410          * \r
 411          * @param input\r
 412          * @return\r
 413          * @throws IOException\r
 414          * @throws UnknownFileFormatException\r
 415          */\r
 416         static List<MultiAnnotatedSequence<DisemblResultAnnot>> readDisembl(\r
 417                         final InputStream input) throws IOException,\r
 418                         UnknownFileFormatException {\r
 419                 Scanner scan = new Scanner(input);\r
 420                 scan.useDelimiter("# RESIDUE COILS REM465 HOTLOOPS\n");\r
 421                 if (!scan.hasNext()) {\r
 422                         throw new UnknownFileFormatException(\r
 423                                         "In Disembl score format each seqeunce score is expected to start from the line: "\r
 424                                                         + "'# RESIDUE COILS REM465 HOTLOOPS\\n'."\r
 425                                                         + " No such line was found!");\r
 426                 }\r
 427 \r
 428                 List<MultiAnnotatedSequence<DisemblResultAnnot>> results = new ArrayList<MultiAnnotatedSequence<DisemblResultAnnot>>();\r
 429                 int seqCounter = 0;\r
 430                 while (scan.hasNext()) {\r
 431                         seqCounter++;\r
 432                         String singleSeq = scan.next();\r
 433                         Scanner scansingle = new Scanner(singleSeq);\r
 434                         StringBuffer seqbuffer = new StringBuffer();\r
 435                         ArrayList<Float> coils = new ArrayList<Float>();\r
 436                         ArrayList<Float> rem = new ArrayList<Float>();\r
 437                         ArrayList<Float> hotloops = new ArrayList<Float>();\r
 438 \r
 439                         MultiAnnotatedSequence<DisemblResultAnnot> disemblRes = new MultiAnnotatedSequence<DisemblResultAnnot>(\r
 440                                         DisemblResultAnnot.class);\r
 441 \r
 442                         while (scansingle.hasNextLine()) {\r
 443                                 String valueLine = scansingle.nextLine();\r
 444                                 Scanner values = new Scanner(valueLine);\r
 445                                 seqbuffer.append(values.next());\r
 446                                 coils.add(values.nextFloat());\r
 447                                 rem.add(values.nextFloat());\r
 448                                 hotloops.add(values.nextFloat());\r
 449                                 values.close();\r
 450                         }\r
 451                         disemblRes.addAnnotation(DisemblResultAnnot.COILS, coils);\r
 452                         disemblRes.addAnnotation(DisemblResultAnnot.REM465, rem);\r
 453                         disemblRes.addAnnotation(DisemblResultAnnot.HOTLOOPS, hotloops);\r
 454                         // TODO\r
 455                         // disemblRes.sequence = seqbuffer.toString();\r
 456                         scansingle.close();\r
 457                         results.add(disemblRes);\r
 458                 }\r
 459 \r
 460                 input.close();\r
 461                 return results;\r
 462         }\r
 463 \r
 464         /**\r
 465          * Read AACon result with no alignment files. This method leaves incoming\r
 466          * the InputStream results open!\r
 467          * \r
 468          * @param results\r
 469          *            output file of AAConservation\r
 470          * @return Map with keys {@link Method} -> float[]\r
 471          */\r
 472         public static HashSet<Score> readAAConResults(InputStream results) {\r
 473                 if (results == null) {\r
 474                         throw new NullPointerException(\r
 475                                         "InputStream with results must be provided");\r
 476                 }\r
 477                 HashSet<Score> annotations = new HashSet<Score>();\r
 478                 Scanner sc = new Scanner(results);\r
 479                 sc.useDelimiter("#");\r
 480                 while (sc.hasNext()) {\r
 481                         String line = sc.next();\r
 482                         int spacePos = line.indexOf(" ");\r
 483                         assert spacePos > 0 : "Space is expected as delimited between method "\r
 484                                         + "name and values!";\r
 485                         String methodLine = line.substring(0, spacePos);\r
 486                         Method method = Method.getMethod(methodLine);\r
 487                         assert method != null : "Method " + methodLine\r
 488                                         + " is not recognized! ";\r
 489                         Scanner valuesScanner = new Scanner(line.substring(spacePos));\r
 490                         ArrayList<Float> values = new ArrayList<Float>();\r
 491                         while (valuesScanner.hasNextDouble()) {\r
 492                                 Double value = valuesScanner.nextDouble();\r
 493                                 values.add(value.floatValue());\r
 494                         }\r
 495                         annotations.add(new Score(method, values));\r
 496                 }\r
 497                 return annotations;\r
 498         }\r
 499 \r
 500 }\r