The phosphate chemistry of nucleoside triphosphates (NTPs) is at the heart of crucial biological processes including the cell’s energy storage, signalling and copying genetic information. Computational modelling valuably expands the spatiotemporal resolution of experimental structural biology, providing a key contribution to the synergistic understanding of such complex processes. However, our preliminary results on phosphatase reactions shed light on the qualitative inaccuracy of the state-of-the-art methodology capable of free energy bio-simulations. Thus, the candidate will be trained to extend the host’s uniquely accurate and efficient electronic structure approaches to the scale of enzyme simulations, producing thermodynamic properties comparable to experiments. The candidate brings complementary computational biochemistry experience and reaction mechanistic knowledge to this collaboration, demonstrated in his successful computational-experimental phosphate catalytic enzyme studies. Our concerted developments will combine methods from the forefront of quantum chemistry, high-performance computing, rare event molecular dynamics, and biochemical modelling into a widely applicable open-access protocol, with a predictive power currently limited to main-group homogeneous catalysis modelling. The reliability of the reaction mechanisms predicted by the developed method will be validated against kinetic experiments on wild-type and mutant Ras GTPases, connected to ca. one third of the cancer cell lines. Then, we will utilise the established protocol to understand the (regular and faulty) binding and incorporation of deoxy-NTPs and nucleoside analogue drugs during mitochondrial DNA polymerisation to a yet unknown atomistic detail. These results will provide a foundation for a comprehensive kinetic model for DNA polymerase fidelity and for the future rational optimisation of more selective and less toxic antiretrovirals.