The fundamental interest is to develop scalable mathematical and numerical analysis techniques that increase the predictive fidelity of biomolecular network models, by incorporating physico-chemical constraints, motivated by optimality principles. Their applied interest is in the aetiopathogenesis and amelioration of neurodegenrative disease, especially Parkinson’s disease. Model predictions are used for optimal experimental design and compared with quantitative experimental data, including that obtained from microfluidic cell cultures of dopaminergic neurons, derived from normal and Parkinsonian human subjects using stem cell biology techniques.