The protein of interest (PDB) that I downloaded from the Protein data bank website, comes as a biological assembly ( not as an asymmetric unit). Should I delete the repeated amino acid chain or I can use it directly for Docking?
It depends on what question you want to address and what is the active oligomeric unit of that protein is biological functional. Also, is your active site between different oligomer unit of protein or is it on single protein?
You can download either. Whether you need the biological assembly, part of it or only a single subunit of a multimeric enzyme for docking depends on where the suspected active site is located. If you know it is entirely contained within one subunit, you can get away with using only that subunit, if it is formed by the interface between two subunits, you may need to use more than one subunit. In any case, you should check that the docking solutions do not clash with other subunits in the full assembly.
It depends on the location of the binding site. If your consider ligand, binds between two oligomers, you must use the symmetric form of PDB structure in the docking process. if you don't have any information on the binding site, you can optimize the docking process, predict more possible binding pocket on your PDB structure by available tools.