News

Domain-based Structural Alignments

05/22

The sequence diagram for PDB ID 3BMV shows the corresponding UniProtKB sequence, the SEQRES and ATOM records, and the various annotations that are available.

(Click image to enlarge)
The sequence diagram for PDB ID 3BMV shows the corresponding UniProtKB sequence, the SEQRES and ATOM records, and the various annotations that are available.

Access pre-calculated systematic structure comparisons for all proteins from a Structure Summary's 3D Similarity tab.

In the example of the 3D Similarity tab for 3BMV, selecting

(Click image to enlarge)
In the example of the 3D Similarity tab for 3BMV, selecting "view" for the 7th ranked domain PDP:3DHUAa launches the structure alignment view for the alpha amylase domains of 3BMV and 3DHU.  Selecting the link for PDP:3DHUAa under the column Domain 2 returns the 3D Similarity tab for entry 3DHU.

To provide more accurate results, the latest version of the 3D Similarity tab uses domain-based protein structure alignments instead of chain-based alignments.

For an example, see the 3D Similarity tab for glucanotransferase (PDB ID 3BMV).

An image of the sequence highlights how the residues listed in the sequence (SEQRES) and in the atom records (ATOM) map onto the relevant parts of the UniProtKB sequence, along with annotations from DSSP, SCOP, PDP and Pfam.

Domains are also highlighted in a table that displays the most important calculated results and scores. Domains can be selected from the pull down menu above the table, or by clicking on a domain in the sequence image.

The table can be sorted and filtered, and offers links to a 3D structure alignment in Jmol (from the results column) and to information about similar domains. In the example of the 3D Similarity tab for 3BMV, selecting "view" for the 7th ranked domain PDP:3DHUAa launches the structure alignment view for the alpha amylase domains of 3BMV and 3DHU.  Selecting the link for PDP:3DHUAa under the column Domain 2 returns the 3D Similarity tab for entry 3DHU.

The calculation of the domain-split representative is an extension of our sequence clustering approach. To remove redundancy, we start with a 40% sequence identity clustering procedure, and select a representative chain from each sequence cluster. If the representative chain contains multiple domains, each is included. SCOP 1.75 domain assignments are used when available. Otherwise, the assignments are computed using ProteinDomainParser.

News Index