This returns the percentage sequence identity between this sequence and seq, which is defined as the number of identical aligned residues divided by the length of the shorter sequence.