Overbinding, qualitative and quantitative changes caused by simple Na+ and K+ ions in polyelectrolyte simulations: comparison of force fields with and without NBFIX and ECC corrections.

Overbinding of ions is a common and well-known problem in classical molecular dynamics (MD) simulations. One of its main causes is the absence of electronic polarizability in the force fields. The current approaches for minimizing overbinding typically either retain the original charges and use an ad hoc re-adjustment of the Lennard-Jones parameters as done in the non-bonded fix (NBFIX) approach or re-scaling of charges using a theoretical framework. The goal in the latter is to include shielding produced by the missing electronic polarizability as done in the electronic continuum correction (ECC) approach. NBFIX and ECC are the most common corrections and we compare their performance to the default parameterizations provided by five different commonly used biomolecular force fields, OPLS-AA/L, CHARMM27, CHARMM36m, CHARMM22* and AMBER99SB-ILDN. As test systems, we use poly-α,L-glutamic (PGA) and poly-α,L-aspartic (PASA) amino acids molecules in explicit water together with Na+ and K+ counterions. We demonstrate that the different force fields yield results that are not only quantitatively but also qualitatively different. The resulting structures of the macroions depend strongly on the model for ions. NBFIX corrections alleviate the problem of overbinding, resulting in extended peptides. The ECC corrections depend non-trivially on the original underlying model and despite being based on a theoretical framework, cannot always solve the problem.