Abstract

Principal Component Analysis (PCA) and Artificial Neural Network (ANN) were used to analyze the relationship between the structure and the activities of a series of nine biphenylphenyl methanone derivatives against Mycobacterium tuberculosis in vitro. Both PCA and ANN were able to classify these derivatives in two categories: low active and highly active compounds. Empirical and theoretical descriptors were used in the classification process. The descriptors selected by PCA indicated that the reactivity plays an important role in the determination of antimycobacterial activity of biphenylphenyl methanone derivatives (BPM). The BPM showed a moderate activity against the M. tuberculosis strain tested with the exception of chloride-, bromide- and nitroderivatives (when X = Cl, Br, NO2) which were the most actives against M. tuberculosis in vitro among all the methanones studied.