Identification of key residues in proteins by using their physical characters

Key residues in proteins are important to their stability, folding, and functions. They usually are highly conserved and can be identified by sequence or structure alignments. However, these methods can only determine the locations of key residues in sequences and structures and give less information about their physical characters. In this paper, we try to identify key residues by analyzing their inter-residue interactions. The model we study is the protein domain from transducin. We show that the usual Gaussian network analysis and distance-based contact analysis have difficulty identifying the key residues in this protein, but the contact energies can do it well. We find that most key residues can be located by the lowest contact energies. This enables us to predict and analyze the key residues in other proteins. Our results suggest that contact energy analysis may provide an alternative approach to investigating the folding and stability of proteins.