RESUMO
Calcium silicate (C3S) cements are available in kits that do not account for patients' specific needs or clinicians' preferences regarding setting time, radiopacity, mechanical, and handling properties. Moreover, slight variations in powder components and liquid content affect cement's properties and bioactivity. Unfortunately, it is virtually impossible to optimize several cement properties simultaneously via the traditional "one variable at a time" strategy, as inputs often induce trade-offs in properties (e.g., a higher water-to-powder ratio [W/P] increases flowability but decreases mechanical properties). Herein, we used Taguchi's methods and genetic algorithms (GAs) to simultaneously analyze the effect of multiple inputs (e.g., powder composition, radiopacifier concentration, and W/P) on setting time, pH, flowability, diametral tensile strength, and radiopacity, as well as prescribe recipes to produce cements with predicted properties. The properties of cements designed with GAs were experimentally tested, and the results matched the predictions. Finally, we show that the cements increased the genetic expression of odonto/osteogenic genes, alkaline phosphatase activity, and mineralization potential of dental pulp stem cells. Hence, GAs can produce cements with tailor-made properties and differentiation potential for personalized endodontic treatment.