The analysis of superfamily variability in the CATH database discussed above is presented as a significant update to the Dictionary of Homologous Superfamilies (DHS) web-resource (Bray et al. 2001 ref). The DHS also provides both structural and functional annotations of domains within each H-level (superfamily) in CATH v2.5.1.
Multiple alignments of all S35Reps across a superfamily are presented in the form of a 2DSEC diagram, alongside co-ordinate data of the superposed structures in PDB format. These alignments are performed using the residue-based CORA algorithm (Orengo et al. 1999 ref). Sequence representations of the alignments are available to download in FASTA format. A picture of the alignment is also shown, with residues in each domain coloured according to ligand binding and residue type. An EquivSEC plot is also shown that describes the variability in orientation between equivalent secondary structures.
As shown by the analysis in the corresponding paper, a number of superfamilies are highly structurally diverse. Consequently, automatic multiple structural alignments across the whole superfamily can be become unreliable and are often subjective. For this reason, the DHS also provides alignments within structurally similar subgroups (SSGs). Domains are first subjected to multi-linkage clustering using a SSAP score cutoff of 85. The subsequent SSG clusters are then aligned using CORA. Even within a S35 sequence family, sequence alignment methods can go awry. As these often contain domains of similar function, multiple structure alignments are also provided for the user.
Information and links to other functional databases (ENZYME, GO, KEGG, COG, SWISSPROT) are also included for all domains in the superfamily. These were created by using BLAST to scan the associated PDB chain against UniProt sequences in the BIOMAP database (ref?). Only 95% sequence identity hits to gene entries were taken as genuine matches.
Reference
Bray JE, Todd AE, Pearl FM, Thornton JM, Orengo CA. (2000)
The CATH Dictionary of Homologous Superfamilies (DHS): a consensus approach for identifying distant structural homologues. Protein Eng. 13(3):153-65.
