Classifying a Protein in the CATH Database of Domain Structures

Orengo, C.A.; Martin, Andrew C.R.; Hutchinson, Gillian; Jones, Susan R.; Jones, Dave; Michie, Alex; Swindells, Mark B.; Thornton, Janet M.

doi:10.1107/s0907444998007501

Cited by 30 publications

(20 citation statements)

References 28 publications

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…The opposite effect is observed in the case of CATH fold 1.20.120, which covers proteins such as 1oedB (acetylcholine receptor, beta chain) and 1fftC (ubiquinol oxidase) with only a weak structural similarity (Dali Z ‐score of 8.1, SSAP score of 66, and SSAP RMSD as high as 7.63 Å) that are in fact not fulfilling the requirements of a common CATH classification (required SSAP score > 70). Although these proteins end up in the same CATH fold due to the single‐linkage clustering approach used during classification,50 SCOP places them in different folds (f.25 andf.36), which have homogenous functional GO assignments39 and high average Z ‐scores (24.3 for fold f.25 and 19.4 for fold f.36).…”

Section: Resultsmentioning

confidence: 99%

Current status of membrane protein structure classification

2010

View full text Add to dashboard Cite

For over 2 decades, continuous efforts to organize the jungle of available protein structures have been underway. Although a number of discrepancies between different classification approaches for soluble proteins have been reported, the classification of membrane proteins has so far not been comparatively studied because of the limited amount of available structural data. Here, we present an analysis of alpha-helical membrane protein classification in the SCOP and CATH databases. In the current set of 63 alpha-helical membrane protein chains having between 1 and 13 transmembrane helices, we observed a number of differently classified proteins both regarding their domain and fold assignment. The majority of all discrepancies affect single transmembrane helix, two helix hairpin, and four helix bundle domains, while domains with more than five helices are mostly classified consistently between SCOP and CATH. It thus appears that the structural constraints imposed by the lipid bilayer complicate the classification of membrane proteins with only few membrane-spanning regions. This problem seems to be specific for membrane proteins as soluble four helix bundles, not restrained by the membrane, are more consistently classified by SCOP and CATH. Our findings indicate that the structural space of small membrane helix bundles is highly continuous such that even minor differences in individual classification procedures may lead to a significantly different classification. Membrane proteins with few helices and limited structural diversity only seem to be reasonably classifiable if the definition of a fold is adapted to include more fine-grained structural features such as helix-helix interactions and reentrant regions.

show abstract

Section: Resultsmentioning

confidence: 99%

Current status of membrane protein structure classification

2010

View full text Add to dashboard Cite

show abstract

“…This is an enhancement of a first version that comprised only 116 PDB codes (Lonquety et al, 2008a). The distribution in the Topology level of CATH (Orengo et al, 1998), corresponding to the fold level of SCOP (Murzin et al, 1995), is given in Table 1, with the exception of 18 PDB codes unavailable in CATH. The coverage of the folds is rather low (7.1% of the total Topology level of CATH are present in our dataset), because our attention was focused on constituting families able to evidence topohydrophobic positions, already demonstrated to fit to MIR (Papandreou et al, 2004), from the multiple alignments within sequences of the same Homology level.…”

Section: Datasetmentioning

confidence: 99%

Prediction of Stability upon Point Mutation in the Context of the Folding Nucleus

Lonquety

Chomilier²,

Παπανδρεου³

et al. 2010

OMICS: A Journal of Integrative Biology

View full text Add to dashboard Cite

Proteins come in all shapes and sizes. Although it is possible to predict with reasonable success their structure from their sequence, the process of folding a chain of amino acids into its tertiary structure remains partially understood. This article addresses several characteristics pertaining to protein folding. The development of the Most Interacting Residues (MIR) algorithm, which dynamically simulates the early folding events, permits a reasonable ab initio prediction of the deeply buried critical residues involved in the formation of the protein core. The analysis of MIR positions with respect to protein 3D topology, in particular, to fragments called Tightened End Fragments (TEF) that might be good candidate for autonomous folding units, suggests that they are also essential for defining core stability. To validate this hypothesis, this study measures the sensitivity of MIR residues to point mutations. It is performed on a set of 385 proteins from a database that contains stability data calculated with five different algorithms. Tools have been developed to help the analysis and a consensus of the five methods is proposed. It results that positions predicted both as a MIR and a minimum of stability for the consensus are good candidates for the folding nucleus, and consequently their mutations may be hazardous.

show abstract

“…These databases build their families starting from experimentally determined structural domains. SUPERFAMILY is based on the SCOP classification (21), and Gene3D is based on the CATH classification (22). SUPERFAMILY had a human sequence coverage of 72% and residue coverage of 41%, while Gene3D covered 69% of human sequences and 35% of human residues.…”

Section: Resultsmentioning

confidence: 99%

The challenge of increasing Pfam coverage of the human proteome

Mistry

Coggill²,

Eberhardt³

et al. 2013

Database

View full text Add to dashboard Cite

It is a worthy goal to completely characterize all human proteins in terms of their domains. Here, using the Pfam database, we asked how far we have progressed in this endeavour. Ninety per cent of proteins in the human proteome matched at least one of 5494 manually curated Pfam-A families. In contrast, human residue coverage by Pfam-A families was <45%, with 9418 automatically generated Pfam-B families adding a further 10%. Even after excluding predicted signal peptide regions and short regions (<50 consecutive residues) unlikely to harbour new families, for ∼38% of the human protein residues, there was no information in Pfam about conservation and evolutionary relationship with other protein regions. This uncovered portion of the human proteome was found to be distributed over almost 25 000 distinct protein regions. Comparison with proteins in the UniProtKB database suggested that the human regions that exhibited similarity to thousands of other sequences were often either divergent elements or N- or C-terminal extensions of existing families. Thirty-four per cent of regions, on the other hand, matched fewer than 100 sequences in UniProtKB. Most of these did not appear to share any relationship with existing Pfam-A families, suggesting that thousands of new families would need to be generated to cover them. Also, these latter regions were particularly rich in amino acid compositional bias such as the one associated with intrinsic disorder. This could represent a significant obstacle toward their inclusion into new Pfam families. Based on these observations, a major focus for increasing Pfam coverage of the human proteome will be to improve the definition of existing families. New families will also be built, prioritizing those that have been experimentally functionally characterized.Database URL: http://pfam.sanger.ac.uk/

show abstract

Classifying a Protein in the CATH Database of Domain Structures

Cited by 30 publications

References 28 publications

Current status of membrane protein structure classification

Current status of membrane protein structure classification

Prediction of Stability upon Point Mutation in the Context of the Folding Nucleus

The challenge of increasing Pfam coverage of the human proteome

Contact Info

Product

Resources

About