EN FR
EN FR


Section: Application Domains

Protein comparison

Comparing protein is important for understanding their evolutionary relationships and for predicting their structures and their functions. While annotating functions for new proteins, such as those solved in structural genomics projects, protein structural alignment methods may be able to identify functionally related proteins when the sequence identity between a given query protein and the related proteins are low (i.e. lower than 20%). Moreover, protein comparison allows for solving the so-called protein family identification problem. Given an unclassified protein structure (query), the comparison of protein structures can be used for assigning a score measuring the "similarity" between the query and the proteins belonging to a set of families. Based on this score, the query is assigned to one of the families of the set. The knowledge acquired by performing such analyses can then be exploited in methods for protein structure prediction that are based on a homology modeling approach.