The score is defined by number of amino acids of the original gene found in the duplicated gene divided by the number of amino acids of the original gene. This means that mismatches are counted, but gaps are not counted.