I'd be very very worried about a protein model that just came out alphafold to be used to come up with drug targets without at least some molecular dynamics simulation of the solution within which that protein works with. There may be other glycans, ligands, or wholly different metabollites that could affect this binding.