2
$\begingroup$

I tried to find a homolog that has a PDB structure available so I can use this PDB file for comparative modeling. I ran a BLAST search but none of the search results seemed to have a known structure. Is there any way to find the results I want?

$\endgroup$
1
$\begingroup$

If you are using NCBI Blast and blastp against the pdb database gives you nothing, you can do a psiblast against the normal refseq database, do a few rounds, save the PSSM matrix and submit a new job with the PSSM matrix but against the PDB database.

You might need to limit the species of you get lots of too close homologues —say blasting E. coli protein for which you'd want to exclude Enterobacteriales.

$\endgroup$
1
$\begingroup$

To expand the first answer by @MatteoFurla: re-run the P-blast and under database option, which is set by default to non-redundant database (nr) select Protein databank (PDB), the third option from the bottom of the menu. All the Blast results will be from PDB and thus all results will have associated structures. Also note that the Genbank codes will be PDB codes.

If you have zero hits then consider PSI-Blast, how I suspect you are running P-Blast via the nr database.

$\endgroup$
1
$\begingroup$

You could use HMM-based comparison, for example with HHpred. This is generally the best way to find structural templates because structure is more conserved than sequence, and comparing HMMs will turn up hits of the same structure but minimal sequence overlap.

If HHpred turns up nothing you are in the region of de novo modelling, which is a different problem. However, it has advanced in recent years to the point where it may be useful to you. For example, you could try our method DMPfold.

$\endgroup$

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.