ABSTRACT
This chapter looks at a wide variety of schemes for measuring similarity and conservation between amino acids in aligned protein sequences. Molecular evolution offers two contrasting perspectives embodied in the neutralist and selectionist models. The chapter considers a neutralist model because that model is by far the more commonly applied in structure and sequence bioinformatics. The identity matrix for scoring amino acid similarity is clearly naive. Genetic code scoring considers the minimum number of DNA/RNA base changes that would be required to interconvert the codons of one amino acid to those of the other amino acid. The basic unit of molecular evolution expressed in a Mutation Data Matrix (MDM) is the Point Accepted Mutation (PAM). Empirical studies have shown that the BLOSUM matrices provide a more useful measure of amino acid similarity for the purposes of comparing amino acid sequences or databank searching.
