Abstarct: The rock mass deformation modulus (D_f) is a highly influential parameter in analytical and numerical deformability analyses. The in-situ tests are the best approach to determine D_f, but there is the challenge of cost and time-consuming. In the last few decades, by conducting in-situ tests, the databaxse of in-site tests was gradually increased, and the geomechanical information like in-situ stress improved. Moreover, the effect of factors such as RQD, RMR, Vp,… on the D_f has been studied experimentally. However, the in situ stress effects haven’t been considered, and its impact on the D_f has not been investigated comprehensively. The mentioned challenges prompted geoengineers to estimate the D_f baxsed on available information, focusing on rock engineering classifications. However, the lack of theoretical support for the input parameters was the main challenge in the prediction process in the majority of empirical models have presented, yet. Therefore, despite the simplicity of the presented models, their estimated D_f suffer from a lack of reliability. Ordinarily, a conservative approach leads to choosing the lowest D_f among the range of estimated deformation modulus. In this research, first, the permutation of theoretical-baxsed influential parameters on D_f is introduced. Second, to overcome the structural limitations of conventional regression methods (linear, exponential, power, polynomial), the multigene genetic programming (MGP) approach as a soft computing method was applied to present the first empirical equation with theoretical support that considered in-situ stress and practical issues together. The MGP approach provides a mathematical relationship with an acceptable fitness function ("R" ^"2" "≥0.6" ) that is verified by comparing with existing empirical correlations baxsed on the same input data. Third, back analysis carried out by numerical modelling (3DEC) revealed the considerable role of in situ stress on predicted D_f. The new horizon to predict D_f is supported by confined Young's modulus and scaled joint stiffness that participated in the presented MGP model. Applying the MGP model baxsed on analytical input parameters leads to considerable savings in terms of time and cost.