Functional families and structural clusters in CATH

Functional Families in CATH (CATH-FunFams)

The homologous superfamilies in CATH can often be functionally and structurally diverse even though they share a conserved structural core. They have therefore been sub-classified into functional families (FunFams) using a subclassification protocol called FunFHMMER. This method is currently based on identifying sequence patterns that discriminate between sets of relatives having different functions. Sequence data is used as the structural data from the PDB is considerably more sparse than the available sequence data.

Relatives within CATH FunFams are likely to share highly similar structures and functions. Thus, FunFams help to understand the mechanisms of functional divergence in a superfamily during evolution. Functional family classification is a useful means to improve the functional annotation of sequences of uncharacterised protein domains, assigned to an annotated functional family within the superfamily. 

Structural clusters

Protein domain structures within a superfamily are compared using an in-house method, Sequential structure alignment program (SSAP), and then clustered if they are structurally similar. Two types of structural clusters are generated by this method:

1. Moderate similarity clusters (< 9 Å RMSD).
2. High similarity clusters (< 5 Å RMSD). These clusters of structurally similar domains are also known as structurally similar groups (SSGs).

These structural clusters are useful for understanding the structural diversity of a superfamily.

To learn more about protein domains, family and classification check out the EMBL-EBI online tutorial: Protein classification: An introduction to EMBL-EBI resources