A crystal structure by definition is experimental, derived by interpreting X-ray diffraction on a protein crystal. If you design a structure in silico, what you have is a model

To ab initio design a protein sequence that will bind a given epitope in the antigen is very hard. in contast to DNA/RNA, where you have a simple algorithm to design an antisense sequence by the rules of base complementarity, with proteins there is no simple rule of amino acid complementarity where amino acid A binds amino acid B. The interaction depends on how the proteins folds into a 3D shape, and binding depends on both shape and electrostatic/chemical complementarity.

While it is easy to homology model the antibody structure as a whole, the complementary determining loops that make up the paratope are highly divergent between different antibodies, especially CDR H3, which in most antibody/antigen complexes contributes most to the binding interface.

I agree with Annemarie Honegger. I think it is difficult to build the model under such circumstances. If you can obtain the experimental crystal structure, it will be helpful for subsequent experiments and analysis, and the results will be more reliable. If you don’t have the resources or equipment, you can solve the problem by working with a trusted partner. Creative Biostructure is an expert in the field of crystal structure determination of biomacromolecules. I think it is great. Maybe you could try it.