2.1 Translating Computational Chemistry to Predictive Toxicology
Computational chemistry has made impressive strides over the past two decades. We can now use it to study organic and enzymatic reactions; assess viability of molecular mechanisms; propose novel inhibitors of biological targets in computerized drug discovery; and even attempt de novo design of macromolecules with novel functionalities. To this end, substantial research efforts have been devoted in recent years to adapting and applying methodology from computational chemistry to predictive toxicology. These efforts have yielded powerful new descriptors of molecular events that can be applied to many types of predictive statistical models.
The descriptors we obtain from computational models are molecular properties and the changes that result in the course of a (bio)chemical event. They are influenced by model size (the extent of system representation), level of theory (the rigor of the method), and by system dynamics (evolution of the system to mimic real-life conditions). ...