Fig. 11.

Detailed performance of ForceGen (-pquant) all atom RMSD compared with results from other methods on the Prime MCS Set. Compared with all four other methods, ForceGen’s performance improvement was statistically significant (paired t-test, p less than ${10}^{-4}$ vs. Prime-MCS and MacroModel, ${10}^{-7}$ vs. MOE, and ${10}^{-15}$ vs. MD)