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;
+ /*
+ * an arbitrary score to assign for identity of an unknown symbol
+ * (this is the value on the diagonal in the * column of the NCBI matrix)
+ * (though a case could be made for using the minimum diagonal value)
+ */
+ private static final int UNKNOWN_IDENTITY_SCORE = 1;
+
+ /*
+ * 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)
/*
* 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;
*/
private boolean peptide;
+ private float minValue;
+
+ private float maxValue;
+
/**
* Constructor given a name, symbol alphabet, and matrix of scores for pairs
* of symbols. The matrix should be square and of the same size as the
symbolIndex = buildSymbolIndex(alphabet);
+ findMinMax();
+
/*
* crude heuristic for now...
*/
}
/**
+ * Record the minimum and maximum score values
+ */
+ protected void findMinMax()
+ {
+ float min = Float.MAX_VALUE;
+ float max = -Float.MAX_VALUE;
+ if (matrix != null)
+ {
+ for (float[] row : matrix)
+ {
+ if (row != null)
+ {
+ for (float f : row)
+ {
+ min = Math.min(min, f);
+ max = Math.max(max, f);
+ }
+ }
+ }
+ }
+ minValue = min;
+ maxValue = max;
+ }
+
+ /**
* Returns an array A where A[i] is the position in the alphabet array of the
* character whose value is i. For example if the alphabet is { 'A', 'D', 'X'
* } then A['D'] = A[68] = 1.
* 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;
}
/**
- * Returns the pairwise score for substituting c with d, or zero if c or d is
- * an unscored or unexpected character
+ * 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. If either c or d is
+ * an unexpected character, returns 1 for identity (c == d), else the minimum
+ * score value in the matrix.
*/
@Override
public float getPairwiseScore(char c, char d)
{
return matrix[cIndex][dIndex];
}
- return 0;
+
+ /*
+ * one or both symbols not found in the matrix
+ * currently scoring as 1 (for identity) or the minimum
+ * matrix score value (otherwise)
+ * (a case could be made for using minimum row/column value instead)
+ */
+ return c == d ? UNKNOWN_IDENTITY_SCORE : getMinimumScore();
}
/**
else
{
sb.append("ScoreMatrix ").append(getName()).append("\n");
- sb.append(symbols).append("\n");
}
for (char sym : symbols)
{
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);
*
* @return
*/
- public String getSymbols()
+ String getSymbols()
{
return new String(symbols);
}
+
+ public void setDescription(String desc)
+ {
+ description = desc;
+ }
+
+ public float getMinimumScore()
+ {
+ return minValue;
+ }
+
+ public float getMaximumScore()
+ {
+ return maxValue;
+ }
}