2 * Jalview - A Sequence Alignment Editor and Viewer ($$Version-Rel$$)
3 * Copyright (C) $$Year-Rel$$ The Jalview Authors
5 * This file is part of Jalview.
7 * Jalview is free software: you can redistribute it and/or
8 * modify it under the terms of the GNU General Public License
9 * as published by the Free Software Foundation, either version 3
10 * of the License, or (at your option) any later version.
12 * Jalview is distributed in the hope that it will be useful, but
13 * WITHOUT ANY WARRANTY; without even the implied warranty
14 * of MERCHANTABILITY or FITNESS FOR A PARTICULAR
15 * PURPOSE. See the GNU General Public License for more details.
17 * You should have received a copy of the GNU General Public License
18 * along with Jalview. If not, see <http://www.gnu.org/licenses/>.
19 * The Jalview Authors are detailed in the 'AUTHORS' file.
23 import jalview.analysis.scoremodels.ScoreMatrix;
24 import jalview.analysis.scoremodels.ScoreModels;
25 import jalview.datamodel.SequenceI;
27 import java.io.IOException;
28 import java.util.StringTokenizer;
31 * A class that can parse a file containing a substitution matrix and register
32 * it for use in Jalview
34 * Accepts 'NCBI' format (e.g.
35 * https://www.ncbi.nlm.nih.gov/Class/FieldGuide/BLOSUM62.txt), with the
36 * addition of a header line to provide a matrix name, e.g.
39 * ScoreMatrix BLOSUM62
42 * Also accepts 'AAindex' format (as described at
43 * http://www.genome.jp/aaindex/aaindex_help.html) with the minimum data
47 * H accession number (used as score matrix identifier in Jalview)
48 * D description (used for tooltip in Jalview)
50 * and the substitution scores
53 public class ScoreMatrixFile extends AlignFile
54 implements AlignmentFileReaderI
56 // first non-comment line identifier - also checked in IdentifyFile
57 public static final String SCOREMATRIX = "SCOREMATRIX";
59 private static final String DELIMITERS = " ,\t";
61 private static final String COMMENT_CHAR = "#";
63 private String matrixName;
66 * aaindex format has scores for diagonal and below only
68 boolean isLowerDiagonalOnly;
71 * ncbi format has symbols as first column on score rows
73 boolean hasGuideColumn;
81 public ScoreMatrixFile(FileParse source) throws IOException
87 public String print(SequenceI[] sqs, boolean jvsuffix)
93 * Parses the score matrix file, and if successful registers the matrix so it
94 * will be shown in Jalview menus. This method is not thread-safe (a separate
95 * instance of this class should be used by each thread).
98 public void parse() throws IOException
100 ScoreMatrix sm = parseMatrix();
102 ScoreModels.getInstance().registerScoreModel(sm);
106 * Parses the score matrix file and constructs a ScoreMatrix object. If an
107 * error is found in parsing, it is thrown as FileFormatException. Any
108 * warnings are written to syserr.
111 * @throws IOException
113 public ScoreMatrix parseMatrix() throws IOException
115 ScoreMatrix sm = null;
118 char[] alphabet = null;
119 float[][] scores = null;
124 isLowerDiagonalOnly = false;
126 while ((data = nextLine()) != null)
130 if (data.startsWith(COMMENT_CHAR) || data.length() == 0)
134 if (data.toUpperCase().startsWith(SCOREMATRIX))
137 * Parse name from ScoreMatrix <name>
138 * we allow any delimiter after ScoreMatrix then take the rest of the line
142 throw new FileFormatException(
143 "Error: 'ScoreMatrix' repeated in file at line "
146 StringTokenizer nameLine = new StringTokenizer(data, DELIMITERS);
147 if (nameLine.countTokens() < 2)
149 err = "Format error: expected 'ScoreMatrix <name>', found '"
150 + data + "' at line " + lineNo;
151 throw new FileFormatException(err);
153 nameLine.nextToken(); // 'ScoreMatrix'
154 name = nameLine.nextToken(); // next field
155 name = data.substring(1).substring(data.substring(1).indexOf(name));
158 else if (data.startsWith("H ") && name == null)
163 return parseAAIndexFormat(lineNo, data);
165 else if (name == null)
167 err = "Format error: 'ScoreMatrix <name>' should be the first non-comment line";
168 throw new FileFormatException(err);
172 * next non-comment line after ScoreMatrix should be the
173 * column header line with the alphabet of scored symbols
175 if (alphabet == null)
177 StringTokenizer columnHeadings = new StringTokenizer(data,
179 size = columnHeadings.countTokens();
180 alphabet = new char[size];
182 while (columnHeadings.hasMoreTokens())
184 alphabet[col++] = columnHeadings.nextToken().charAt(0);
186 scores = new float[size][];
191 * too much information
195 err = "Unexpected extra input line in score model file: '" + data
197 throw new FileFormatException(err);
200 parseValues(data, lineNo, scores, row, alphabet);
205 * out of data - check we found enough
210 "Expected %d rows of score data in score matrix but only found %d",
212 throw new FileFormatException(err);
216 * If we get here, then name, alphabet and scores have been parsed successfully
218 sm = new ScoreMatrix(name, alphabet, scores);
225 * Parse input as AAIndex format, starting from the header line with the
231 * @throws IOException
233 protected ScoreMatrix parseAAIndexFormat(int lineNo, String data)
236 String name = data.substring(2).trim();
237 String description = null;
239 float[][] scores = null;
240 char[] alphabet = null;
244 while ((data = nextLine()) != null)
248 if (skipAAindexLine(data))
252 if (data.startsWith("D "))
254 description = data.substring(2).trim();
256 else if (data.startsWith("M "))
258 alphabet = parseAAindexRowsColumns(lineNo, data);
259 size = alphabet.length;
260 scores = new float[size][size];
262 else if (scores == null)
264 throw new FileFormatException(
265 "No alphabet specified in matrix file");
267 else if (row >= size)
269 throw new FileFormatException("Too many data rows in matrix file");
273 parseValues(data, lineNo, scores, row, alphabet);
278 ScoreMatrix sm = new ScoreMatrix(name, description, alphabet, scores);
285 * Parse one row of score values, delimited by whitespace or commas. The line
286 * may optionally include the symbol from which the scores are defined. Values
287 * may be present for all columns, or only up to the diagonal (in which case
288 * upper diagonal values are set symmetrically).
291 * the line to be parsed
294 * the score matrix to add data to
296 * the row number / alphabet index position
300 * if invalid, or too few, or too many values
302 protected void parseValues(String data, int lineNo, float[][] scores,
303 int row, char[] alphabet) throws FileFormatException
306 int size = alphabet.length;
307 StringTokenizer scoreLine = new StringTokenizer(data, DELIMITERS);
309 int tokenCount = scoreLine.countTokens();
312 * inspect first row to see if it includes the symbol in the first column,
313 * and to see if it is lower diagonal values only (i.e. just one score)
317 if (data.startsWith(String.valueOf(alphabet[0])))
319 hasGuideColumn = true;
321 if (tokenCount == (hasGuideColumn ? 2 : 1))
323 isLowerDiagonalOnly = true;
330 * check 'guide' symbol is the row'th letter of the alphabet
332 String symbol = scoreLine.nextToken();
333 if (symbol.length() > 1 || symbol.charAt(0) != alphabet[row])
336 "Error parsing score matrix at line %d, expected '%s' but found '%s'",
337 lineNo, alphabet[row], symbol);
338 throw new FileFormatException(err);
340 tokenCount = scoreLine.countTokens(); // excluding guide symbol
344 * check the right number of values (lower diagonal or full format)
346 if (isLowerDiagonalOnly && tokenCount != row + 1)
349 "Expected %d scores at line %d: '%s' but found %d", row + 1,
350 lineNo, data, tokenCount);
351 throw new FileFormatException(err);
354 if (!isLowerDiagonalOnly && tokenCount != size)
357 "Expected %d scores at line %d: '%s' but found %d", size,
358 lineNo, data, scoreLine.countTokens());
359 throw new FileFormatException(err);
363 * parse and set the values, setting the symmetrical value
364 * as well if lower diagonal format data
366 scores[row] = new float[size];
369 while (scoreLine.hasMoreTokens())
373 value = scoreLine.nextToken();
374 scores[row][col] = Float.valueOf(value);
375 if (isLowerDiagonalOnly)
377 scores[col][row] = scores[row][col];
380 } catch (NumberFormatException e)
382 err = String.format("Invalid score value '%s' at line %d column %d",
384 throw new FileFormatException(err);
390 * Parse the line in an aaindex file that looks like
393 * M rows = ARNDCQEGHILKMFPSTWYV, cols = ARNDCQEGHILKMFPSTWYV
396 * rejecting it if rows and cols do not match. Returns the string of
397 * characters in the row/cols alphabet.
402 * @throws FileFormatException
404 protected char[] parseAAindexRowsColumns(int lineNo, String data)
405 throws FileFormatException
407 String err = "Unexpected aaIndex score matrix data at line " + lineNo
412 String[] toks = data.split(",");
413 String rowsAlphabet = toks[0].split("=")[1].trim();
414 String colsAlphabet = toks[1].split("=")[1].trim();
415 if (!rowsAlphabet.equals(colsAlphabet))
417 throw new FileFormatException("rows != cols");
419 return rowsAlphabet.toCharArray();
420 } catch (Throwable t)
422 throw new FileFormatException(err + " " + t.getMessage());
427 * Answers true if line is one we are not interested in from AAindex format
433 protected boolean skipAAindexLine(String data)
435 if (data.startsWith(COMMENT_CHAR) || data.length() == 0)
439 if (data.startsWith("*") || data.startsWith("R ")
440 || data.startsWith("A ") || data.startsWith("T ")
441 || data.startsWith("J ") || data.startsWith("//"))
448 public String getMatrixName()