Benchmarking Nonpolarizable Mg2+ Parameters Reveals Trade-Offs in RNA Conformational Sampling

In molecular dynamics simulations of RNA, Mg2+ ion force field parameters remain highly sensitive to the choice of water model and nucleic acid force field. The choice of Mg2+ force field parameters requires intentional thought, and although recent trends in nucleic acid force field development increasingly use the OPC water model, the compatibility of Mg2+ parameters with the OL3/OPC combination is insufficiently characterized. Here, we benchmark a panel of Mg2+ force fields: first in water boxes containing OPC with a single Mg2+ ion, and then in simulations with the Varkud satellite ribozyme stem loop V (SLV) as a model RNA, using the OL3 force field with LJbb corrections (OL3-LJbb). While most Mg2+ models are designed to reproduce the experimental ordering of hydration shells, they still differ substantially in diffusion coefficients and hydration free energies, reflecting the known trade-off between reproducing hydration thermodynamics and ion-oxygen distances in divalent-cation parametrization. These differences translate directly to RNA simulations, where Mg2+ models with slower diffusion and overly favorable hydration free energies promote long time-scale RNA binding events (“chelation”) and disrupt Mg2+-driven conformational changes. Together, these results highlight the need to validate divalent ion parameters specifically within the water model and RNA force field environment used for production-phase RNA molecular simulations, as well as demonstrate the Varkud satellite ribozyme as a sensitive benchmark for Mg2+/RNA/solvent force field compatibility.