I would like to calculate the free energy using MMPBSA in different segments of the trajectory. These segments correspond, mostly, to the dominant conformers. Considering that these segments include several frames of other conformers, is it correct to perform an average or if a weighted average should be performed?