I don't know about  antigen-antibody modelling. The Schrodinger suite should have dedicated tools to accomplish such kind of tasks but it is a commercial software and, in my personal opinion, I don't trust with it.

Otherwise MODELLER remains my gold standard to build homology models.

Thanks Dario for your quick response.

Hi,

The first thing you should think is do these missing residues have a particular structure helix, etc are they important for the protein function

If those residues are missing in the PDB file, this can arise from several reasons, among which :

- Those residues needed to be deleted because the full-length protein wasn't soluble/expressed. etc.

According to me modeller would be a best option. You can try SPDBV  it automatically generates missing residues in the structure, you can try it. Hope it helps

Regards

You may try this one

https://salilab.org/modeller/

Thanks Danish and Arafat.

would you all like to tell me please which is the most good template because when i performed modelling using modeller it give me so many PDB IDs that can use as template including my target PDB ( 4I3S) that have 100% similarity and zero e-value as well. so can i use my target PDB as template?

Hi,

Yes you can try modelling with your template, i would also suggest you to try modelling using PHYRE2 server and check the result. Or else simply open your structure in SPDBV. 

Regards.

Alrigh Danish. I will try this.

Usually I don't trust with embedded options for template searching. Usually I trust with the results from psi-blast, Same thing for alignment, in such case I use DALI or MUSTANG for a preliminary structural alignment of the templates, then I fit the query sequence by comparing the predicted secondary structure profile (using PSIPRED for example) with the ones of the templates (obtained with DSSP). I use Modeller only for the final model building.

Thank yo so much Dario. It's good and i definitely  want to compare the result by using these software which is more reliable.