Mechano-bactericidal effects exhibited by specific nano-patterns have brought in the prospect of developing sustainable antibacterial materials. Contrary to the standard practices of administrating anti-bacterial agents or chemical surface functionalization, nano-patterns manage to inactivate a wide variety of bacteria species with no risk of toxicity, antibiotic resistance or replenishment. Herein, the experimental data on the bactericidal effect of nano-patterns were collected to develop in-silico models for identifying the impact of individual geometrical features. An artificial neural network was developed considering the three prevalent species of Escherichia coli, Pseudomonas aeruginosa, and Staphylococcus aureus. The roles of individual geometrical features were analyzed and comprehensive parametric and sensitivity analyses were performed to determine the most favorable range for each parameter against different species. Geometrical features that would demonstrate bactericidal effects simultaneously against all the three studied species were identified. The efficient geometrical parameters, obtained from the artificial neural network analysis, were then used to develop a series of finite element models to simulate the physical interaction between the bacteria and the nano-patterns that result in inactivation. The obtained results can pave the way for unlocking the role of geometrical features towards optimized development of artificial materials with sustainable intrinsic antibacterial characteristics.