Purpose: To predict activity and reveal the pharmacophore (Pha) with certain electronic and topological characteristics for a series of 37 molecules of 4,4′-dihydroxydiphenylmethane, using 4D QSAR (four dimensional Quantitative-Structure Activity Relationships) model.

Methods: We used a computational method called molecular conformer electron topological (MCET) for this study. The quality of Pha and the corresponding quantitative model of activity was validated (and deemed acceptable) by an independent test set of 7 additional analogs with known experimental activities out of 30 molecules of the training set.

Results: The resulting MCET method demonstrated a high statistical capacity for predicting the activity of the molecules under consideration (R²= 0.703 and Q² = 0.573).

Conclusion: The model is based on pure computational methods (electronic structure calculations and matrix comparisons) and provides the correct solution within the assumptions of the method, experimental uncertainty, and computational approximations. A different procedure from other QSAR approaches was used to elucidate the interactions between the conformers of the ligand and the target protein.