src/jalview/io/TCoffeeScoreFile.java

   1 package jalview.io;
   2
   3 import jalview.analysis.SequenceIdMatcher;
   4 import jalview.datamodel.AlignmentAnnotation;
   5 import jalview.datamodel.AlignmentI;
   6 import jalview.datamodel.Annotation;
   7 import jalview.datamodel.SequenceI;
   8
   9 import java.awt.Color;
  10 import java.io.BufferedReader;
  11 import java.io.File;
  12 import java.io.FileNotFoundException;
  13 import java.io.FileReader;
  14 import java.io.IOException;
  15 import java.io.Reader;
  16 import java.util.ArrayList;
  17 import java.util.HashMap;
  18 import java.util.LinkedHashMap;
  19 import java.util.List;
  20 import java.util.Map;
  21
  22 /**
  23  * A file parse for T-Coffee score ascii format. This file contains the alignment consensus
  24  * for each resude in any sequence.
  25  * <p>
  26  * This file is procuded by <code>t_coffee</code> providing the option
  27  * <code>-output=score_ascii </code> to the program command line
  28  *
  29  * An example file is the following
  30  *
  31  * <pre>
  32  * T-COFFEE, Version_9.02.r1228 (2012-02-16 18:15:12 - Revision 1228 - Build 336)
  33  * Cedric Notredame
  34  * CPU TIME:0 sec.
  35  * SCORE=90
  36  * *
  37  *  BAD AVG GOOD
  38  * *
  39  * 1PHT   :  89
  40  * 1BB9   :  90
  41  * 1UHC   :  94
  42  * 1YCS   :  94
  43  * 1OOT   :  93
  44  * 1ABO   :  94
  45  * 1FYN   :  94
  46  * 1QCF   :  94
  47  * cons   :  90
  48  *
  49  * 1PHT   999999999999999999999999998762112222543211112134
  50  * 1BB9   99999999999999999999999999987-------4322----2234
  51  * 1UHC   99999999999999999999999999987-------5321----2246
  52  * 1YCS   99999999999999999999999999986-------4321----1-35
  53  * 1OOT   999999999999999999999999999861-------3------1135
  54  * 1ABO   99999999999999999999999999986-------422-------34
  55  * 1FYN   99999999999999999999999999985-------32--------35
  56  * 1QCF   99999999999999999999999999974-------2---------24
  57  * cons   999999999999999999999999999851000110321100001134
  58  *
  59  *
  60  * 1PHT   ----------5666642367889999999999889
  61  * 1BB9   1111111111676653-355679999999999889
  62  * 1UHC   ----------788774--66789999999999889
  63  * 1YCS   ----------78777--356789999999999889
  64  * 1OOT   ----------78877--356789999999997-67
  65  * 1ABO   ----------687774--56779999999999889
  66  * 1FYN   ----------6888842356789999999999889
  67  * 1QCF   ----------6878742356789999999999889
  68  * cons   00100000006877641356789999999999889
  69  * </pre>
  70  *
  71  *
  72  * @author Paolo Di Tommaso
  73  *
  74  */
  75 public class TCoffeeScoreFile {
  76
  77         /** The {@link Header} structure holder */
  78         Header header;
  79
  80         /**
  81          * Holds the consensues values for each sequences. It uses a LinkedHashMap to maintaint the
  82          * insertion order.
  83          */
  84         LinkedHashMap<String,StringBuilder> scores = new LinkedHashMap<String,StringBuilder>();
  85
  86         Integer fWidth;
  87
  88         /**
  89          * Parse the specified file.
  90          *
  91          * @param file The file to be parsed
  92          */
  93         public static TCoffeeScoreFile load(File file) {
  94                 try {
  95                         return load(new FileReader(file));
  96                 }
  97                 catch (FileNotFoundException e) {
  98                         throw new RuntimeException(e);
  99                 }
 100         }
 101
 102         /**
 103          * Parse the provided reader for the T-Coffee scores file format
 104          *
 105          * @param reader
 106          */
 107         public static TCoffeeScoreFile load(Reader reader) {
 108
 109                 try {
 110                         BufferedReader in = (BufferedReader) (reader instanceof BufferedReader ? reader : new BufferedReader(reader));
 111                         TCoffeeScoreFile result = new TCoffeeScoreFile();
 112                         result.doParsing(in);
 113                         return result.header != null && result.scores != null ? result : null;
 114                 }
 115                 catch( Exception e) {
 116                         throw new RuntimeException(e);
 117                 }
 118         }
 119
 120         /**
 121          * @return The 'height' of the score matrix i.e. the numbers of score rows that should matches
 122          * the number of sequences in the alignment
 123          */
 124         public int getHeight() {
 125                 // the last entry will always be the 'global' alingment consensus scores, so it is removed
 126                 // from the 'height' count to make this value compatible with the number of sequences in the MSA
 127                 return scores != null && scores.size() > 0 ? scores.size()-1 : 0;
 128         }
 129
 130         /**
 131          * @return The 'width' of the score matrix i.e. the number of columns.
 132          * Since teh score value are supposd to be calculated for an 'aligned' MSA, all the entries
 133          * have to have the same width.
 134          */
 135         public int getWidth() {
 136                 return fWidth != null ? fWidth : 0;
 137         }
 138
 139         /**
 140          * The default constructor is marked as {@code protected} since this class is meant to created
 141          * through the {@link #load(File)} or {@link #load(Reader)} factory methods
 142          */
 143         protected TCoffeeScoreFile() { }
 144
 145         /**
 146          * Get the string of score values for the specified seqeunce ID.
 147          * @param id The sequence ID
 148          * @return The scores as a string of values e.g. {@code 99999987-------432}.
 149          *      It return an empty string when the specified ID is missing.
 150          */
 151         public String getScoresFor( String id ) {
 152                 return scores.containsKey(id) ? scores.get(id).toString() : "";
 153         }
 154
 155         /**
 156          * @return The list of score string as a {@link List} object, in the same ordeer of the insertion i.e. in the MSA
 157          */
 158         public List<String> getScoresList() {
 159                 List<String> result = new ArrayList<String>( scores.size() );
 160                 for( Map.Entry<String,StringBuilder> it : scores.entrySet() ) {
 161                         result.add(it.getValue().toString());
 162                 }
 163
 164                 return result;
 165         }
 166
 167         /**
 168          * @return The parsed score values a matrix of bytes
 169          */
 170         public byte[][] getScoresArray() {
 171                 byte[][] result = new byte[ scores.size() ][];
 172
 173                 int rowCount = 0;
 174                 for( Map.Entry<String,StringBuilder> it : scores.entrySet() ) {
 175                         String line = it.getValue().toString();
 176                         byte[] seqValues = new byte[ line.length() ];
 177                         for( int j=0, c=line.length(); j<c; j++ ) {
 178
 179                                 byte val = (byte)(line.charAt(j) - '0');
 180
 181                                 seqValues[j] = ( val >= 0 && val <= 9 ) ? val : -1;
 182                         }
 183
 184                         result[rowCount++] = seqValues;
 185                 }
 186
 187                 return result;
 188         }
 189
 190
 191         private void doParsing(BufferedReader in) throws IOException {
 192
 193                 /*
 194                  * read the header
 195                  */
 196                 header = readHeader(in);
 197
 198                 if( header == null ) { return; }
 199
 200
 201                 /*
 202                  * initilize the structure
 203                  */
 204                 for( Map.Entry<String,Integer> entry : header.scores.entrySet() ) {
 205                         scores.put( entry.getKey(), new StringBuilder());
 206                 }
 207
 208                 /*
 209                  * go with the reading
 210                  */
 211                 Block block;
 212                 while( (block = readBlock(in, header.scores.size())) != null  ) {
 213
 214                         /*
 215                          * append sequences read in the block
 216                          */
 217                         for( Map.Entry<String,String> entry : block.items.entrySet() ) {
 218                                 StringBuilder scoreStringBuilder = scores.get(entry.getKey());
 219                                 if( scoreStringBuilder == null ) {
 220                                         throw new RuntimeException(String.format("Invalid T-Coffee score file: Sequence ID '%s' is not declared in header section", entry.getKey()));
 221                                 }
 222
 223                                 scoreStringBuilder.append( entry.getValue() );
 224                         }
 225                 }
 226
 227                 /*
 228                  * verify that all rows have the same width
 229                  */
 230                 for( StringBuilder str : scores.values() ) {
 231                         if( fWidth == null ) {
 232                                 fWidth = str.length();
 233                         }
 234                         else if( fWidth != str.length() ) {
 235                                 throw new RuntimeException("Invalid T-Coffee score file: All the score sequences must have the same length");
 236                         }
 237                 }
 238
 239
 240
 241         }
 242
 243
 244         static int parseInt( String str ) {
 245                 try {
 246                         return Integer.parseInt(str);
 247                 }
 248                 catch( NumberFormatException e ) {
 249                         // TODO report a warning ?
 250                         return 0;
 251                 }
 252         }
 253
 254         /**
 255          * Reaad the header section in the T-Coffee score file format
 256          *
 257          * @param reader The scores reader
 258          * @return The parser {@link Header} instance
 259          * @throws RuntimeException when the header is not in the expected format
 260          */
 261         static Header readHeader(BufferedReader reader) {
 262
 263                 Header result = null;
 264                 try {
 265                         result = new Header();
 266                         result.head = reader.readLine();
 267
 268                         String line;
 269
 270                         while( (line = reader.readLine()) != null ) {
 271                                 if( line.startsWith("SCORE=")) {
 272                                         result.score = parseInt( line.substring(6).trim() );
 273                                         break;
 274                                 }
 275                         }
 276
 277                         if( (line=reader.readLine())==null || !"*".equals(line.trim())) return null;
 278                         if( (line=reader.readLine())==null || !"BAD AVG GOOD".equals(line.trim())) return null;
 279                         if( (line=reader.readLine())==null || !"*".equals(line.trim())) return null;
 280
 281                         /*
 282                          * now are expected a list if sequences ID up to the first blank line
 283                          */
 284                         while( (line=reader.readLine()) != null ) {
 285                                 if( "".equals(line) ) {
 286                                         break;
 287                                 }
 288
 289                                 int p = line.indexOf(":");
 290                                 if( p == -1 ) {
 291                                         // TODO report a warning
 292                                         continue;
 293                                 }
 294
 295                                 String id = line.substring(0,p).trim();
 296                                 int val = parseInt(line.substring(p+1).trim());
 297                                 if( "".equals(id) ) {
 298                                         // TODO report warning
 299                                         continue;
 300                                 }
 301
 302                                 result.scores.put(id,val);
 303                         }
 304
 305                 }
 306                 catch( IOException e ) {
 307                         throw new RuntimeException("Cannot parse T-Coffee score ascii file", e);
 308                 }
 309
 310                 return result;
 311         }
 312
 313         /**
 314          * Read a scores block ihe provided stream.
 315          *
 316          * @param reader The stream to parse
 317          * @param size The expected number of the sequence to be read
 318          * @return The {@link Block} instance read or {link null} null if the end of file has reached.
 319          * @throws IOException Something went wrong on the 'wire'
 320          */
 321         static Block readBlock( BufferedReader reader, int size ) throws IOException {
 322                 Block result = new Block(size);
 323                 String line;
 324
 325                 /*
 326                  * read blank lines (eventually)
 327                  */
 328                 while( (line=reader.readLine()) != null && "".equals(line.trim())) {
 329                         // consume blank lines
 330                 }
 331
 332                 if( line == null ) return null;
 333
 334                 /*
 335                  * read the scores block
 336                  */
 337                 do {
 338                         if( "".equals(line.trim()) ) {
 339                                 // terminated
 340                                 break;
 341                         }
 342
 343                         // split the line on the first blank
 344                         // the first part have to contain the sequence id
 345                         // theramining part are the scores values
 346                         int p = line.indexOf(" ");
 347                         if( p == -1 ) {
 348                                 //TODO This is an unexpected condition, log a warning or throw an exception ?
 349                                 continue;
 350                         }
 351
 352                         String id = line.substring(0,p).trim();
 353                         String val = line.substring(p+1).trim();
 354
 355                         result.items.put(id, val);
 356
 357                 } while( (line = reader.readLine()) != null );
 358
 359
 360                 return result;
 361         }
 362
 363         /*
 364          * The score file header
 365          */
 366         static class Header {
 367                 String head;
 368                 int score;
 369
 370                 LinkedHashMap<String,Integer> scores = new LinkedHashMap<String,Integer>();
 371
 372                 public int getScoreAvg() { return score; }
 373
 374                 public int getScoreFor( String ID ) {
 375
 376                         return scores.containsKey(ID) ? scores.get(ID) : -1;
 377
 378                 }
 379         }
 380
 381         /*
 382          * Hold a single block values block in the score file
 383          */
 384         static class Block {
 385                 int size;
 386                 Map<String,String> items;
 387
 388                 public Block( int size ) {
 389                         this.size = size;
 390                         this.items = new HashMap<String,String>(size);
 391                 }
 392
 393                 String getScoresFor( String id ) {
 394                         return items.get(id);
 395                 }
 396
 397                 String getConsensus() {
 398                         return items.get("cons");
 399                 }
 400         }
 401         /**
 402          * TCOFFEE score colourscheme
 403          */
 404         static final Color[] colors = {
 405                         new Color( 102, 102, 255 ),     // #6666FF
 406                         new Color( 0, 255, 0),          // #00FF00
 407                         new Color( 102, 255, 0),        // #66FF00
 408                         new Color( 204, 255, 0),        // #CCFF00
 409                         new Color( 255, 255, 0),        // #FFFF00
 410                         new Color( 255, 204, 0),        // #FFCC00
 411                         new Color( 255, 153, 0),        // #FF9900
 412                         new Color( 255, 102, 0),        // #FF6600
 413                         new Color( 255, 51, 0),         // #FF3300
 414                         new Color( 255, 34, 0)          // #FF2000
 415                 };
 416         public final static String TCOFFEE_SCORE="TCoffeeScore";
 417         /**
 418          * generate annotation for this TCoffee score set on the given alignment
 419          * @param al alignment to annotate
 420          * @param matchids if true, annotate sequences based on matching sequence names
 421          * @return true if alignment annotation was modified, false otherwise.
 422          */
 423         public boolean annotateAlignment(AlignmentI al, boolean matchids)
 424         {
 425           boolean added=false;
 426           int i=0;
 427           SequenceIdMatcher sidmatcher = new SequenceIdMatcher(al.getSequencesArray());
 428           byte[][] scoreMatrix=getScoresArray();
 429           // for 2.8 - we locate any existing TCoffee annotation and remove it first before adding this.
 430           for (Map.Entry<String,StringBuilder> id:scores.entrySet())
 431           {
 432             byte[] srow=scoreMatrix[i];
 433             SequenceI s;
 434             if (matchids)
 435             {
 436               s=sidmatcher.findIdMatch(id.getKey());
 437             } else {
 438               s=al.getSequenceAt(i);
 439             }
 440             i++;
 441             if (s==null && i!=scores.size() && !id.getKey().equals("cons"))
 442             {
 443               System.err.println("No "+(matchids ? "match ":" sequences left ")+" for TCoffee score set : "+id.getKey());
 444               continue;
 445             }
 446             int jSize=al.getWidth()< srow.length ? al.getWidth() : srow.length;
 447             Annotation[] annotations=new Annotation[al.getWidth()];
 448             for (int j=0;j<jSize;j++) {
 449               byte val = srow[j];
 450               annotations[j]=new Annotation(s==null ? ""+val:null,s==null ? ""+val:null,(char) val,val*1f,val >= 0 && val < colors.length ? colors[val] : Color.white);
 451             }
 452             AlignmentAnnotation aa=null;
 453             if (s!=null)
 454             {
 455               // TODO - set per sequence score
 456               aa=new AlignmentAnnotation(TCOFFEE_SCORE, "Score for "+id.getKey(), annotations);
 457
 458               aa.setSequenceRef(s);
 459               aa.visible=false;
 460               aa.belowAlignment=false;
 461             } else {
 462               aa=new AlignmentAnnotation("T-COFFEE", "TCoffee column reliability score", annotations);
 463               aa.belowAlignment=true;
 464               aa.visible=true;
 465
 466             }
 467             al.addAnnotation(aa);
 468             added=true;
 469           }
 470           return added;
 471         }
 472
 473
 474 }