[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: [modeller_usage] Problem regarding 3D structure prediction



Hi
When you said that psiBLAST did not return results, did you run it against pdb or against nr? I recently found a case where a blast against pdb return nothing, but a psiblast against nr return results and some of them had structures. We were able to made a very good model (we end up having the xtal structure so we could check) using hmmer's jackhmmer against uniprot and using the best result with known structure in the final alignment (-A option) and the % of identity was below 30.

I hope this helps.

OCS


Oscar Conchillo Solé
Group of Computational Biology and Proteomics
IBB Data Center Manager and Linux Sysadmin
Institut de Biotecnologia I Biomedicina (UAB)
mail: 
telf: 0034 93581 4431; 0034 93586 8939


On 07/21/2014 07:23 PM, Modeller Caretaker wrote:
On 7/20/14, 4:51 AM, Vishal Nemaysh wrote:
I have faced a problem during 3D structure prediction of my target
protein of interest. The length of the protein sequence is (1106 amino
acids). Sir i have tried to find out some templates for covering the
whole query sequence during BLAST or psiBLAST. But there is no any
single template or multiple templates are found to occupied this whole
query sequence.

In that case your only options are to try more sensitive template selection methods (e.g. profile based) or to build models of only part of the sequence.

    Ben Webb, Modeller Caretaker