Development of a simple QSAR model for reliable evaluation of acetylcholinesterase inhibitor potency

With the aging of the western population, more and more people are affected by the neurodegenerative Alzheimer's and Parkinson's disease. Inhibitors of acetylcholinesterase (AChE) have proven to be effective in the treatment of disease symptoms. We report the QSAR regression model for the estimation of potency of a set of 94 structurally diverse compounds (oximes, N-hydroxyiminoacetamides, 4-aminoquinolines and flavonoids) to inhibit AChE, pK_i (AChE). The model is based on three simple descriptors: the valence molecular connectivity index of the zero-order, ⁰χ^v, combined with the number of 10-membered rings (nR10) and number of hydroxyl groups in a molecule (nOH). QSAR model yielded r = 0.947, S.E. = 0.51 and S.E._cv= 0.53; the range of pK_i (exp) = 6.03. It showed its stability when the set of 94 compounds was enlarged, comprising 184 compounds in total (r = 0.886, S.E. = 0.85 and S.E._cv = 0.88; the range of pK_i (exp) = 10.21), resulting in regression parameters which were similar, although only for ⁰χ^v coefficients within the limits of S.E. (0.167(13) and 0.172(16) for the set with 94 and 184 compounds, respectively. The predictive power of the model was shown by the prediction of pK_i values for 61 randomly chosen compounds (S.E._test = 0.86) from the calibration model made on the other 123 compounds (S.E. = 0.85), all taken from the pool of 184 compounds. QSAR descriptors ⁰χ^v, nR10 and nOH were well chosen for describing the interactions of the AChE active site (amino acid interaction) with ligands through the estimation of the inhibitory potency.