Comment on “Hydrogen bonding and stacking interactions of nucleic acid base pairs: A density-functional-theory treatment” [J. Chem. Phys. 114, 5149 (2001)]

A density-functional-theory (DFT) based treatment for calculating the interaction energies of nucleic acid base pairs proposed recently by Elstner et al. [J. Chem. Phys. 114, 5149 (2001)] is criticized. It is shown that dispersion energy that they evaluated amounts to approximately only 30% of the value obtained in rigorous ab initio calculations. The use of the second-order Mo/ller–Plesset perturbation theory results obtained with the 6-31G*(0.25) basis set as a benchmark for validating the DFT results is also criticized. It is suggested that the method of Elstner et al. requires much more stringent testing before it can be deemed to be “very accurate.”