Calculating free energy differences between levels of theory (i.e., \(\Delta A^{low \to high}\)) is integral to performing indirect (S)QM/MM free energy simulations. However, connecting levels of theory via free energy simulations has proved difficult due to (1) bond/angle degrees of freedom, (2) dihedral degrees of freedom, and (3) solvent arrangement differences between levels of theory, largely due to partial charge differences between levels of theory. In order to improve calculation of (S)QM/MM free energy simulations, the free energy simulation community should begin to compare methods based on convergence success relative to overall computational time and resource requirements. We have begun to compile such a dataset by calculating \(\Delta A^{MM \to SCC-DFTB}\) in gas phase for 22 drug-like molecules, as seen in our recent publication, Kearns, et al. 2018, Molecules, Submitted, and we hope that future practitioners will do the same. With this work we hope to provide a standard for comparison for future FES methodologies; additionally, in the near future we hope to continue to add to this dataset including results in more complicated environments such as in solution and in enzyme. All data can be found in our publication and in the accompanying Supporting Information; raw data (such as simulation trajectories and raw data files) can be made available upon request. The purpose of this dataset publication is to make available all starting coordinates, topologies, parameter sets, and input files necessary to replicating the results published in our work.

This research was funded by NIGMS of the National Institutes of Health under award number RO1GM129519. Additionally, FLK acknowledges funding support from NSF's Graduate Research Fellowship Program, project number 3900101301. Finally, FLK, LW, and HLW would like recognize additional computational support was provided, through collaboration with University of South Florida's Computing Cluster (CIRCE) team, via the Major Research Instrumentation Program (MRI-1531590). Finally, SB gratefully acknowledges support for this work from the Austrian Science Fund/FWF (P31024).