How to use python to create a new amino acid substitution scoring matrix (like blosmun62) based on existing multiple amino acid alignment sequences