import java.util.Arrays;
+/**
+ * A class that models a substitution score matrix for any given alphabet of
+ * symbols
+ */
public class ScoreMatrix implements SimilarityScoreModelI,
PairwiseScoreModelI
{
/*
+ * this fields records which gap character (if any) is used in the alphabet;
+ * space, dash or dot are recognised as gap symbols
+ */
+ private char gapCharacter = '0';
+
+ /*
* Jalview 2.10.1 treated gaps as X (peptide) or N (nucleotide)
* for pairwise scoring; 2.10.2 uses gap score (last column) in
* score matrix (JAL-2397)
/*
* the name of the model as shown in menus
+ * each score model in use should have a unique name
*/
private String name;
/*
+ * a description for the model as shown in tooltips
+ */
+ private String description;
+
+ /*
* the characters that the model provides scores for
*/
private char[] symbols;
* of symbols. The matrix should be square and of the same size as the
* alphabet, for example 20x20 for a 20 symbol alphabet.
*
- * @param name
+ * @param theName
* Unique, human readable name for the matrix
* @param alphabet
* the symbols to which scores apply
- * @param matrix
+ * @param values
* Pairwise scores indexed according to the symbol alphabet
*/
- public ScoreMatrix(String name, char[] alphabet, float[][] matrix)
+ public ScoreMatrix(String theName, char[] alphabet, float[][] values)
{
- if (alphabet.length != matrix.length)
+ if (alphabet.length != values.length)
{
throw new IllegalArgumentException(
"score matrix size must match alphabet size");
}
- for (float[] row : matrix)
+ for (float[] row : values)
{
if (row.length != alphabet.length)
{
}
}
- this.matrix = matrix;
- this.name = name;
+ this.matrix = values;
+ this.name = theName;
this.symbols = alphabet;
symbolIndex = buildSymbolIndex(alphabet);
* Mappings are added automatically for lower case symbols (for non case
* sensitive scoring), unless they are explicitly present in the alphabet (are
* scored separately in the score matrix).
+ * <p>
+ * the gap character (space, dash or dot) included in the alphabet (if any) is
+ * recorded in a field
*
* @param alphabet
* @return
*/
- static short[] buildSymbolIndex(char[] alphabet)
+ short[] buildSymbolIndex(char[] alphabet)
{
short[] index = new short[MAX_ASCII + 1];
Arrays.fill(index, UNMAPPED);
short pos = 0;
for (char c : alphabet)
{
+ if (Comparison.isGap(c))
+ {
+ gapCharacter = c;
+ }
+
if (c <= MAX_ASCII)
{
index[c] = pos;
}
@Override
+ public String getDescription()
+ {
+ return description;
+ }
+
+ @Override
public boolean isDNA()
{
return !peptide;
}
/**
+ * Answers the matrix index for the gap character, or -1 if unmapped in the
+ * matrix. Use this method only if using <code>getMatrix</code> in order to
+ * compute scores directly (without symbol lookup) for efficiency.
+ *
+ * @return
+ * @see #getMatrix()
+ */
+ public int getGapIndex()
+ {
+ return getMatrixIndex(gapCharacter);
+ }
+
+ /**
* Returns the pairwise score for substituting c with d, or zero if c or d is
* an unscored or unexpected character
*/
public MatrixI findSimilarities(AlignmentView seqstrings,
SimilarityParamsI options)
{
- char gapChar = scoreGapAsAny ? (seqstrings.isNa() ? 'N' : 'X') : ' ';
+ char gapChar = scoreGapAsAny ? (seqstrings.isNa() ? 'N' : 'X')
+ : Comparison.GAP_DASH;
String[] seqs = seqstrings.getSequenceStrings(gapChar);
return findSimilarities(seqs, options);
}
break;
}
}
- // Change GAP_SPACE to GAP_DASH if we adopt - for gap in matrices
- char c1 = i >= len1 ? Comparison.GAP_SPACE : seq1.charAt(i);
- char c2 = i >= len2 ? Comparison.GAP_SPACE : seq2.charAt(i);
+
+ char c1 = i >= len1 ? gapCharacter : seq1.charAt(i);
+ char c2 = i >= len2 ? gapCharacter : seq2.charAt(i);
boolean gap1 = Comparison.isGap(c1);
boolean gap2 = Comparison.isGap(c2);
/*
* gap-residue: score if options say so
*/
- if (!params.includesGaps())
+ if (!params.includeGaps())
{
continue;
}
{
return new String(symbols);
}
+
+ public void setDescription(String desc)
+ {
+ description = desc;
+ }
}