import java.util.Arrays;
+/**
+ * A class that models a substitution score matrix for any given alphabet of
+ * symbols
+ */
public class ScoreMatrix implements SimilarityScoreModelI,
PairwiseScoreModelI
{
- private static final char GAP_CHARACTER = Comparison.GAP_DASH;
+ /*
+ * 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)
* 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;
}
/**
+ * 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
*/
SimilarityParamsI options)
{
char gapChar = scoreGapAsAny ? (seqstrings.isNa() ? 'N' : 'X')
- : Comparison.GAP_DASH;
+ : gapCharacter;
String[] seqs = seqstrings.getSequenceStrings(gapChar);
return findSimilarities(seqs, options);
}
}
}
- char c1 = i >= len1 ? GAP_CHARACTER : seq1.charAt(i);
- char c2 = i >= len2 ? GAP_CHARACTER : 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);