Predictive folding of RNA structures with molecular dynamics is difficult because of the known inaccuracies of current force fields. However, even testing force fields is a challenge since converged simulations require huge amounts of computer time. We have developed a method based on replica exchange and metadynamics that allows to assess a force field by providing results in a time that is an order of magnitude less than standard methods. Importantly, our calculation not only say when a force field has problems, but also allow those problems to be quantified. We think this will provide a significant boost in the development of new force fields. This work has been done in collaboration with Pavel Banáš and Jiří Šponer.
Results have been published on the Journal of Physical Chemistry Letters.