If you mean the cross-terms for non-bonded interactions, and if you are using simple LJ potentials, a first and good approximation is just to use Lorentz-Berthelot mixing rule:

sig_12 = 0.5 * (sig_11 + sig_22)

eps_12 = sqrt (eps_11 * eps_22).

Dear Dr. Sadeghi

many thanks to your kind response.

My case is associated with organic/mineral interactions and I recently came across a worth reading paper suggesting a procedure for obtaining cross terms parameters in cases identical to me.

Article New Forcefields for Modeling Biomineralization Processes

Have you got experience in deriving cross term? If not, could you please introduce me an expert on this subject?

Many sincere thanks in advance.