You have to do DFT+ NEGF calculations of the electron transport through your molecular junction, and then follow the procedure to get the model parameters (Gamma and so on ) by fitting the TWO-SITE MODEL to ab initio results obtained, as described in the Supplemental Material of the paper you have cited.
I note that ATK is meant to do ab initio part of this calculation, whereas you have to do the fitting of the model yourself. Keep in mind that Gamma also depends on your system of study, meaning that you may need to adopt more complex analytical models for different molecules.
