Does the inclusion of electronic polarisability lead to a better modelling of peptide aggregation?

Simulating the process of amyloid aggregation with atomic detail is a challenging task for various reasons. One of them is that it is difficult to parametrise a force field such that all protein states ranging from the folded through the unfolded to the aggregated state are represented with the same level of accuracy. Here, we test whether the consideration of electronic polarisability improves the description of the different states of Aβ_16-22. Surprisingly, the CHARMM Drude polarisable force field is found to perform worse than its unpolarisable counterpart CHARMM36m. Sources for this failure of the Drude model are discussed.