• Atomistic Simulations
• Force Field
• Machine Learning

Molecular Dynamics (MD) is the prime computational tool for the investigation of basic functional properties of biomolecules in a typical cellular environment, characterized by the presence of ions. The effectiveness of MD investigations depends on the quality of the adopted force field. This thesis work has the purpose to optimize the parameters of nonbonded force field of a chloride ion, Cl-, in soft matter, by means of machine learning techniques, without altering the functional form and the parameters of the force field of the other atoms in the system. The parameters obtained as the main results of this work can be easily integrated into consolidated Molecular Mechanics packages.