Computational Drug Discovery Approach Based on Nuclear Factor-κB Pathway Dynamics

The NF-${\kappa}B$ system of transcription factors plays a crucial role in inflammatory diseases, making it an important drug target. We combined quantitative structure activity relationships for predicting the activity of new compounds and quantitative dynamic models for the NF-${\kappa}B$ network with intracellular concentration models. GFA-MLR QSAR analysis was employed to determine the optimal QSAR equation. To validate the predictability of the $IKK{\beta}$ QSAR model for an external set of inhibitors, a set of ordinary differential equations and mass action kinetics were used for modeling the NF-${\kappa}B$ dynamic system. The reaction parameters were obtained from previously reported research. In the IKKb QSAR model, good cross-validated $q^2$ (0.782) and conventional $r^2$ (0.808) values demonstrated the correlation between the descriptors and each of their activities and reliably predicted the $IKK{\beta}$ activities. Using a developed simulation model of the NF-${\kappa}B$ signaling pathway, we demonstrated differences in $I{\kappa}B$ mRNA expression between normal and different inhibitory states. When the inhibition efficiency increased, inhibitor 1 (PS-1145) led to long-term oscillations. The combined computational modeling and NF-${\kappa}B$ dynamic simulations can be used to understand the inhibition mechanisms and thereby result in the design of mechanism-based inhibitors.