Dear all,

I have generated a database of protein sequences and selected a few sequences from it based on some criteria. Now I would like to model these selected sequences and see if they folded into the native structure of a protein. The protein (native) shares a sequence similarity of 40% with all the sequences in the database.

As I-tasser uses NR database from pdb to generate sequence profile for the target sequence,will it be theoretically correct to use a locally created database of proteins instead of NR database.

Hope I am able to make myself clear. Please ask for more clarity if required and suggest if the idea is feasible or not.

Thank you