Getting discriminant functions of antibacterial activity from physicochemical and topological parameters.

Linear discriminant analysis has been demonstrated to be a very useful tool in the selection and design of new drugs. Up to now we have used it through the search of a topological pattern of activity. In this work our goal is to calculate a complete set of physicochemical parameters using semiempirical (quantum chemical) calculations as well as topological indices (TIs) and try to find out any discriminant function for antibacterial activity through the combined use of both types of descriptors. The physicochemical parameters, such as heat of formation, HOMO, LUMO, dipole moment, polarizability, hyperpolarizability, PM3 generated IR vibrational frequencies, etc., were calculated using PM3 Hamiltonian implemented within the MOPAC97 package. Among the TIs, connectivity as well as topological charge indices stands as the most representatives. The obtained results suggest that one of the maxima and minima vibrational frequencies play an important role in the antibacterial activity. These frequencies are associated with the torsional molecular vibration (N3) and the stretching vibration (N5) of X-H groups (X = C, N, O). Furthermore, the differences between the maxima and minima values showed an even better discriminant ability than the values themselves. The additional use of the topological indices provided a clear improvement in the discriminant function and also provided a straightforward way to predict the values of such frequencies, so that the results can be applied to a large set of compounds searching for new candidates as antibacterials.