src/jalview/io/TCoffeeScoreFile.java

   1 package jalview.io;
   2
   3 import java.io.BufferedReader;
   4 import java.io.File;
   5 import java.io.FileNotFoundException;
   6 import java.io.FileReader;
   7 import java.io.IOException;
   8 import java.io.Reader;
   9 import java.util.ArrayList;
  10 import java.util.HashMap;
  11 import java.util.LinkedHashMap;
  12 import java.util.List;
  13 import java.util.Map;
  14
  15 /**
  16  * A file parse for T-Coffee score ascii format. This file contains the alignment consensus
  17  * for each resude in any sequence.
  18  * <p>
  19  * This file is procuded by <code>t_coffee</code> providing the option
  20  * <code>-output=score_ascii </code> to the program command line
  21  *
  22  * An example file is the following
  23  *
  24  * <pre>
  25  * T-COFFEE, Version_9.02.r1228 (2012-02-16 18:15:12 - Revision 1228 - Build 336)
  26  * Cedric Notredame
  27  * CPU TIME:0 sec.
  28  * SCORE=90
  29  * *
  30  *  BAD AVG GOOD
  31  * *
  32  * 1PHT   :  89
  33  * 1BB9   :  90
  34  * 1UHC   :  94
  35  * 1YCS   :  94
  36  * 1OOT   :  93
  37  * 1ABO   :  94
  38  * 1FYN   :  94
  39  * 1QCF   :  94
  40  * cons   :  90
  41  *
  42  * 1PHT   999999999999999999999999998762112222543211112134
  43  * 1BB9   99999999999999999999999999987-------4322----2234
  44  * 1UHC   99999999999999999999999999987-------5321----2246
  45  * 1YCS   99999999999999999999999999986-------4321----1-35
  46  * 1OOT   999999999999999999999999999861-------3------1135
  47  * 1ABO   99999999999999999999999999986-------422-------34
  48  * 1FYN   99999999999999999999999999985-------32--------35
  49  * 1QCF   99999999999999999999999999974-------2---------24
  50  * cons   999999999999999999999999999851000110321100001134
  51  *
  52  *
  53  * 1PHT   ----------5666642367889999999999889
  54  * 1BB9   1111111111676653-355679999999999889
  55  * 1UHC   ----------788774--66789999999999889
  56  * 1YCS   ----------78777--356789999999999889
  57  * 1OOT   ----------78877--356789999999997-67
  58  * 1ABO   ----------687774--56779999999999889
  59  * 1FYN   ----------6888842356789999999999889
  60  * 1QCF   ----------6878742356789999999999889
  61  * cons   00100000006877641356789999999999889
  62  * </pre>
  63  *
  64  *
  65  * @author Paolo Di Tommaso
  66  *
  67  */
  68 public class TCoffeeScoreFile {
  69
  70         /** The {@link Header} structure holder */
  71         Header header;
  72
  73         /**
  74          * Holds the consensues values for each sequences. It uses a LinkedHashMap to maintaint the
  75          * insertion order.
  76          */
  77         LinkedHashMap<String,StringBuilder> scores = new LinkedHashMap<String,StringBuilder>();
  78
  79
  80         /**
  81          * Parse the specified file.
  82          *
  83          * @param file The file to be parsed
  84          */
  85         public static TCoffeeScoreFile load(File file) {
  86                 try {
  87                         return load(new FileReader(file));
  88                 }
  89                 catch (FileNotFoundException e) {
  90                         throw new RuntimeException(e);
  91                 }
  92         }
  93
  94         /**
  95          * Parse the provided reader for the T-Coffee scores file format
  96          *
  97          * @param reader
  98          */
  99         public static TCoffeeScoreFile load(Reader reader) {
 100
 101                 try {
 102                         BufferedReader in = (BufferedReader) (reader instanceof BufferedReader ? reader : new BufferedReader(reader));
 103                         TCoffeeScoreFile result = new TCoffeeScoreFile();
 104                         result.doParsing(in);
 105                         return result.header != null && result.scores != null ? result : null;
 106                 }
 107                 catch( Exception e) {
 108                         throw new RuntimeException(e);
 109                 }
 110         }
 111
 112         /**
 113          * The default constructor is marked as {@code protected} since this class is meant to created
 114          * through the {@link #load(File)} or {@link #load(Reader)} factory methods
 115          */
 116         protected TCoffeeScoreFile() { }
 117
 118         /**
 119          * Get the string of score values for the specified seqeunce ID.
 120          * @param id The sequence ID
 121          * @return The scores as a string of values e.g. {@code 99999987-------432}.
 122          *      It return an empty string when the specified ID is missing.
 123          */
 124         public String getScoresFor( String id ) {
 125                 return scores.containsKey(id) ? scores.get(id).toString() : "";
 126         }
 127
 128         /**
 129          * @return The list of score string as a {@link List} object, in the same ordeer of the insertion i.e. in the MSA
 130          */
 131         public List<String> getScoresList() {
 132                 List<String> result = new ArrayList<String>( scores.size() );
 133                 for( Map.Entry<String,StringBuilder> it : scores.entrySet() ) {
 134                         result.add(it.getValue().toString());
 135                 }
 136
 137                 return result;
 138         }
 139
 140         /**
 141          * @return The parsed score values a matrix of bytes
 142          */
 143         public byte[][] getScoresArray() {
 144                 byte[][] result = new byte[ scores.size() ][];
 145
 146                 int rowCount = 0;
 147                 for( Map.Entry<String,StringBuilder> it : scores.entrySet() ) {
 148                         String line = it.getValue().toString();
 149                         byte[] seqValues = new byte[ line.length() ];
 150                         for( int j=0, c=line.length(); j<c; j++ ) {
 151
 152                                 byte val = (byte)(line.charAt(j) - '0');
 153
 154                                 seqValues[j] = ( val >= 0 && val <= 9 ) ? val : -1;
 155                         }
 156
 157                         result[rowCount++] = seqValues;
 158                 }
 159
 160                 return result;
 161         }
 162
 163
 164         private void doParsing(BufferedReader in) throws IOException {
 165
 166                 /*
 167                  * read the header
 168                  */
 169                 header = readHeader(in);
 170
 171                 if( header == null ) { return; }
 172
 173
 174                 /*
 175                  * initilize the structure
 176                  */
 177                 for( Map.Entry<String,Integer> entry : header.scores.entrySet() ) {
 178                         scores.put( entry.getKey(), new StringBuilder());
 179                 }
 180
 181                 /*
 182                  * go with the reading
 183                  */
 184                 Block block;
 185                 while( (block = readBlock(in, header.scores.size())) != null  ) {
 186
 187                         /*
 188                          * append sequences read in the block
 189                          */
 190                         for( Map.Entry<String,String> entry : block.items.entrySet() ) {
 191                                 StringBuilder scoreStringBuilder = scores.get(entry.getKey());
 192                                 if( scoreStringBuilder == null ) {
 193                                         throw new RuntimeException(String.format("Invalid T-Coffee score file. Sequence ID '%s' is not declared in header section", entry.getKey()));
 194                                 }
 195
 196                                 scoreStringBuilder.append( entry.getValue() );
 197                         }
 198
 199                 }
 200
 201         }
 202
 203
 204         static int parseInt( String str ) {
 205                 try {
 206                         return Integer.parseInt(str);
 207                 }
 208                 catch( NumberFormatException e ) {
 209                         // TODO report a warning ?
 210                         return 0;
 211                 }
 212         }
 213
 214         /**
 215          * Reaad the header section in the T-Coffee score file format
 216          *
 217          * @param reader The scores reader
 218          * @return The parser {@link Header} instance
 219          * @throws RuntimeException when the header is not in the expected format
 220          */
 221         static Header readHeader(BufferedReader reader) {
 222
 223                 Header result = null;
 224                 try {
 225                         result = new Header();
 226                         result.head = reader.readLine();
 227
 228                         String line;
 229
 230                         while( (line = reader.readLine()) != null ) {
 231                                 if( line.startsWith("SCORE=")) {
 232                                         result.score = parseInt( line.substring(6).trim() );
 233                                         break;
 234                                 }
 235                         }
 236
 237                         if( (line=reader.readLine())==null || !"*".equals(line.trim())) return null;
 238                         if( (line=reader.readLine())==null || !"BAD AVG GOOD".equals(line.trim())) return null;
 239                         if( (line=reader.readLine())==null || !"*".equals(line.trim())) return null;
 240
 241                         /*
 242                          * now are expected a list if sequences ID up to the first blank line
 243                          */
 244                         while( (line=reader.readLine()) != null ) {
 245                                 if( "".equals(line) ) {
 246                                         break;
 247                                 }
 248
 249                                 int p = line.indexOf(":");
 250                                 if( p == -1 ) {
 251                                         // TODO report a warning
 252                                         continue;
 253                                 }
 254
 255                                 String id = line.substring(0,p).trim();
 256                                 int val = parseInt(line.substring(p+1).trim());
 257                                 if( "".equals(id) ) {
 258                                         // TODO report warning
 259                                         continue;
 260                                 }
 261
 262                                 result.scores.put(id,val);
 263                         }
 264
 265                 }
 266                 catch( IOException e ) {
 267                         throw new RuntimeException("Cannot parse T-Coffee score ascii file", e);
 268                 }
 269
 270                 return result;
 271         }
 272
 273         /**
 274          * Read a scores block ihe provided stream.
 275          *
 276          * @param reader The stream to parse
 277          * @param size The expected number of the sequence to be read
 278          * @return The {@link Block} instance read or {link null} null if the end of file has reached.
 279          * @throws IOException Something went wrong on the 'wire'
 280          */
 281         static Block readBlock( BufferedReader reader, int size ) throws IOException {
 282                 Block result = new Block(size);
 283                 String line;
 284
 285                 /*
 286                  * read blank lines (eventually)
 287                  */
 288                 while( (line=reader.readLine()) != null && "".equals(line.trim())) {
 289                         // consume blank lines
 290                 }
 291
 292                 if( line == null ) return null;
 293
 294                 /*
 295                  * read the scores block
 296                  */
 297                 do {
 298                         if( "".equals(line.trim()) ) {
 299                                 // terminated
 300                                 break;
 301                         }
 302
 303                         // split the line on the first blank
 304                         // the first part have to contain the sequence id
 305                         // theramining part are the scores values
 306                         int p = line.indexOf(" ");
 307                         if( p == -1 ) {
 308                                 //TODO This is an unexpected condition, log a warning or throw an exception ?
 309                                 continue;
 310                         }
 311
 312                         String id = line.substring(0,p).trim();
 313                         String val = line.substring(p+1).trim();
 314
 315                         result.items.put(id, val);
 316
 317                 } while( (line = reader.readLine()) != null );
 318
 319
 320                 return result;
 321         }
 322
 323         /*
 324          * The score file header
 325          */
 326         static class Header {
 327                 String head;
 328                 int score;
 329
 330                 LinkedHashMap<String,Integer> scores = new LinkedHashMap<String,Integer>();
 331
 332                 public int getScoreAvg() { return score; }
 333
 334                 public int getScoreFor( String ID ) {
 335
 336                         return scores.containsKey(ID) ? scores.get(ID) : -1;
 337
 338                 }
 339         }
 340
 341         /*
 342          * Hold a single block values block in the score file
 343          */
 344         static class Block {
 345                 int size;
 346                 Map<String,String> items;
 347
 348                 public Block( int size ) {
 349                         this.size = size;
 350                         this.items = new HashMap<String,String>(size);
 351                 }
 352
 353                 String getScoresFor( String id ) {
 354                         return items.get(id);
 355                 }
 356
 357                 String getConsensus() {
 358                         return items.get("cons");
 359                 }
 360         }
 361
 362
 363
 364 }