Stacking interactions are an important contributor to drug binding, as heterocycles are common in drugs. Beyond π-stacking interactions, unconventional stacking interactions of heterocyclic drug fragments with backbone amide bonds, and Arg-Asp salt bridges can also contribute significantly to the enthalpy of drug binding. These interactions can be quite strong, but are not generally well understood. I will present highly-accurate computed interaction energies for model amides, salt bridges, and the aromatic amino acids stacked with heterocyclic rings common in pharmaceuticals. In all cases, trends in binding energies can be described using readily-computed electrostatic properties of the heterocycles. Ultimately these properties are used to explain trends in binding geometries, as well as to generate multi-parameter predictive models for binding energies. These models should prove useful as a tool for scoring these interactions in drug binding sites.