Support exponential sampling from the space of steady states of a metabolic network

My proposal is to extend the Python package dingo with two Markov Chain Monte Carlo sampling methods, which will enable exponential sampling from the set of flux vectors for constraint-based metabolic modelling. The sampling methods - NUTS Reflective Hamiltonian Monte Carlo, Reflective Exact HMC, and Riemannian Hamiltonian Monte Carlo - have already been implemented in C++ and integrated into the volesti codebase. The extended C++ bindings and Python wrappers will enable access to these methods through dingo. This will involve running experiments on benchmark metabolic networks to compare the results obtained from the new random walks and the existing Multiphase Monte Carlo Sampling method based on Billiard walk. The experiments will generate new biological insights by providing an exponential sampling of flux vectors, which will allow for an unbiased characterization of the metabolic capabilities of biochemical networks. The final deliverables of the project will be the extended dingo package, the results of the experiments, and a brief report discussing the findings.