I was trying to calculate the eigenstate of molecularly projected self consistent Hamiltonian (MPSH) for two terminal molecular device in ATK-2016.3. The zero bias calculation can be done as it is given in the tutorials , but finite bias calculation is not there. I am looking for the bias dependent calculation. My question is that for finite bias MPSH calculation would I have to optimize the device geometry first in presence of applied bias? Please suggest me.

Thanks in advance