Protein structure modeling-1


Before beginning any homology modelling, verify that the templates in Blastp (protein blast) are suitable for constructing the model for your query sequence using the pdb as the search set. 
If you obtain greater than 40% sequence similarity, you can proceed with homology modelling. 
If you have Schrodinger, you can use the PRIME homology modelling module in this suite.
Otherwise, I-TASSER, SWISS-MODEL, or Phyre2 servers may be used. 
Additionally, Modeller can be utilized. It is a self-contained programme.
If the sequence similarity of your protein is really low, then try to search with DELTA-BLAST (Domain Enhanced Lookup Time Accelerated BLAST) option in BLASTp. It will yield better homology detection.
Or you can proceed with threading-based or ab-initio modelling. 
Additionally, secondary structure prediction can be performed prior to modelling. This can be accomplished with PSIPRED or JPRED, for example.
There are two forms of modelling: single template modelling and multiple template modelling. Thus, prior to modelling, one must conduct thorough research on the query and templates, as well as the part of the protein that needs to be modelled. Occasionally, modelling the entire protein is unnecessary if the sequence similarity is very low. Therefore, only a required domain can be modelled to begin the study.
Additionally, the Ramachandran plot or the PROCHECK website should be used to verify the protein structure's quality.
Following the model's construction, refine the loops and minimise the structure.
A simulation of molecular dynamics can be used to equilibrate the model and obtain a stable and low-energy state for the model.

Comments

Popular posts from this blog

Calculation of RMSD values between two ligand poses using web servers/tools

AMBER TUTORIAL-2: How to simulate a protein- ligand system: basic steps for md simulation in amber (AMBER 16)

Python script to plot 3D FEL plots