optimization of structure

Dear Lammps users,
Is it really important that the .pdb file(input of Packmol for creating Lammps datafile through Topotools) be optimized properly? I mean, if it has like 20% error in energy compared to the optimized structure on the same basis set in reference, the Lammps simulation will not be correct? Or Lammps minimization will do this work and there is no difference between the correctly optimized .pdb and the .pdb with the error. I am getting parameters and charges from the article, so I think basically the initial structure is just some initial coordination of atoms for parameters to be implemented, am I right?
I am simulating an ion pair, so maybe the distance between oppositely charged ions should be correctly optimized in the first place. yes?
Best regards

There is no simple answer to this Q. If LAMMPS
can minimize and then run dynamics without
things blowing up, then the bad initial structure
is not an issue numerically. But whether
the bad structure locks in some geometry
that is bad for your subsequent analysis
is not a LAMMPS Q.