Binding free energy calculations serve as a physically rigorous approach to predict the potency of compounds and therefore help prioritize the synthesis of potential ligands. Accurate results have been achieved for comparing the binding of related compounds using relative binding free energy calculations (RBFE). The standard approaches, however, require the ligands to share a common scaffold as well as the same binding mode, restricting the domain of applicability of the method. This is a critical limitation since complex modifications of the ligands are very common. Here, we report on our recent work on binding free energy calculations, including work on an alternative method for RBFE, called separated topologies, which was prototyped earlier. Instead of only transforming atoms that vary between two ligands, this approach performs two absolute free energy calculations of the ligands simultaneously in opposite directions. Treating the ligands separately allows researchers to consider diverse ligands. At the same time this approach avoids the need to sample the unbound state of the protein making it more efficient than absolute binding free energy calculations. We tested the approach on binding of related ligands and compared it to traditional RBFE as well as showed its value in complex modifications by predicting ligand transformations that fall outside the scope of standard approaches. Here, we report on our recent work with separated topologies and other techniques and lessons we have learned from those studies.