Domains definition from sequence

Protein domains, the next level up in the hierarchy of protein constitution, are important enough to warrant a separate discussion. Domains can be defined based on function, structure, or sequence characterization in many cases the different approaches are compatible. We naturally adopt a structure-based definition a protein domain is a spatially distinct structure (or structural component) that could conceivably fold and function in isolation [63]. Some proteins consist of a single domain, while others are composed of multiple domains each folded separately from a subsection of the underlying amino acid chain. To this point, our notion of the genotype to phenotype relationship has been protein sequence protein structure. Given the discrete spatial nature of domains, protein SM sequence -> domain would be an equally valid definition. In fact, the notion of a protein fold (as in, "this fold is highly designable") translates naturally to the protein domain concept. 

The aldehyde oxidoreductase from Desulfovibrio gigas shows 52% sequence identity with xanthine oxidase (199, 212) and is, so far, the single representative of the xanthine oxidase family. The 3D structure of MOP was analyzed at 1.8 A resolution in several states oxidized, reduced, desulfo and sulfo forms, and alcohol-bound (200), which has allowed more precise definition of the metal coordination site and contributed to the understanding of its role in catalysis. The overall structure, composed of a single polypeptide of 907 amino acid residues, is organized into four domains two N-terminus smaller domains, which bind the two types of [2Fe-2S] centers and two much larger domains, which harbor the molybdopterin cofactor, deeply buried in the molecule (Fig. 10). The pterin cofactor is present as a cytosine dinucleotide (MCD) and is 15 A away from the molecular surface,... 

---