Defining and predicting structurally conserved regions in protein superfamilies

Studies have revealed that in general, the 3-dimensional structure of a protein is better conserved throughout evolution than its sequence. This phenomenon is perhaps best demonstrated by the prevalence of structurally conserved regions (SCRs) in even highly divergent protein families. Therefore, it is often helpful to characterize these families by their structure rather than their sequence. However, defining SCRs computationally is not straightforward as existing methods tend to underestimate the fraction of SCRs and almost all require the existence of homologous structures to be compared to. The latter situation is particularly troubling, as without ample structural information of protein homologs, it is difficult to develop an understanding of a conserved region for a single protein. Hence, it can be quite practical for a method to identify SCRs using information derived from a single protein.

Defining and predicting structurally conserved regions in protein superfamilies. Huang IK, Pei J, and Grishin NV. Bioinformatics. 2013 Jan 15;29(2):175-181