Résumé: Proteins have been classi .ed into families based on metrics of similarity such as sequence or structural similarity. However, there are signi .cant di .erences in function even within families. Mapping these differences to individual amino- acid residues is typically done by an expert. This is a subjective and non- scalable approach. ConsDi .is an algorithm that automates this process. It is based on a set of parametric rules using amino- acid substitution matrices and a multiple sequence alignment. This allows the automated discovery of candidate residues that may be responsible for critical differences in function, which may then be experimentally verified.

Langue: Anglais