Study of Antidepressant Molecular Structure Leading to Safer Dosing

In this study the structures of 16 molecules with known pKi values for CYP2D6 inhibition were evaluated in silico and quantified using molecular descriptors. The 16 molecules had similar structures, as they were all either SSRIs, SNRIs, or other molecularly similar drugs. The role of the molecular structure of the drugs in predicting the inhibition of the CYP2D6 enzyme was evaluated. This was done using a mathematical model that established a linear relationship between a drug's predicted pKi values and the drug's molecular structure. After measuring the effect that different molecular descriptors had on the accuracy of the predicted pKi values, the math model was limited to 4 descriptors that provided reasonably good fit to predict a drug's pKi inhibition: pKi = 2.56272 + (-0.68072 logS) + (2.49311 PC+) + (-1.48765 SlogP) + (-0.09520 TPSA). Using this model, the predicted pKi values for the 16 molecules in the test set were calculated, and they were cross-validated against the known pKi values for each of the molecules. The model was then used to predict the pKi values for 3 antidepressants whose pKi values for inhibiting 2D6 were unknown. Knowing this is imperative for physicians, as incorrect dosed medications can lead to deadly synergy between various drugs (drug-drug interactions).