On Thu, 22 Aug 2013, Gloria Borgstahl wrote:
We have a protein sequence that probably contains OB folds. What is the
best way to search for the top structural homologs to this sequence in the
pdb? G
Hi Gloria,
If you expect decent sequence simnilarity to one or more proteins in the PDB,
and if you don't have the structure of your protein, a simple FASTA search
should give you the list of hits and alignments quickly. I would do that using
the PDBeXplore FASTA archive browser - http://pdbe.org/fasta
Simply provide your sequence (or UniProt accession) and submit. If there are
hits, they will be presented in a way that allows for rapid further
exploration, e.g. in terms of their CATH classifications, bound ligands or
Pfam families. You can also sort the hits by E-value or %-age identity
(amongst many other things) and get the alignments with your sequence in the
results page.
If you want to explore the interface, try
http://www.ebi.ac.uk/pdbe-srv/PDBeXplore/sequence/?seq=SVRKFTEKHEWVTTENGVGTVGISNFAQEALGDVVYCSLPEVGTKLNKQEEFGALESVKAASELYSPLSGEVTEINKALAENPGLVNKSCYEDGWLIKMTFSNPSELDELMSEEAYEKYIKSIEE&tab=PDB%20entries
(this uses the sequence of PDB entry 3KLR which also has an OB fold, CATH
class 2.40.50).
There are a number of PDBeXplore modules available (http://pdbe.org/explore),
including a CATH-based browser (if you should like to analyse all the
structures that contain an OB fold according to CATH).
If you should have a structure, you can use PDBeFold (SSM;
http://pdbe.org/fold) to quickly find and compare similar structures without
taking sequence into consideration.
--Gerard
