Evolutionarily closely-knit proteins, the homologues, retain similarity at most levels: sequence, structure and function. My laboratory is interested in distant relationships such as ?superfamilies? where small structural variations set in, owing to high evolutionary divergence, that in turn affect precise function. I will present the fascinating problem of how nature provides structural and functional themes amongst protein domains of a superfamily despite poor sequence similarity. I will next present some of our thoughts and analyses on structurally conserved motifs and variant units of protein domain superfamilies revolving around the theme of evolutionary divergence.

Solvent-buried or spatially interacting structurally conserved motifs are good descriptors and signatures of a superfamily. They can effectively drive specificity and sensitivity in sequence searches in genome databases and improve sequence alignment and homology modeling exercises. These motifs can be recognized as regions that undergo permitted amino acid exchanges and substantial equivalence of structural content.

The current analyses suggest that exchange, insertion and deletion of chemical groups happen amongst domains of a protein superfamily in a non-random manner even at very poor sequence identities in protein domain superfamilies.