Thermodynamics-Based Model Construction for the Accurate Prediction of Molecular Properties From Partition Coefficients

Abstract

Developing models for predicting molecular properties of organic compounds is imperative for drug development and environmental safety; however, development of such models that have high predictive power and are independent of the compounds used is challenging. To overcome the challenges, we used a thermodynamics-based theoretical derivation to construct models for accurately predicting molecular properties. The free energy change that determines a property equals the sum of the free energy changes (ΔG_Fs) caused by the factors affecting the property. By developing or selecting molecular descriptors that are directly proportional to ΔG_Fs, we built a general linear free energy relationship (LFER) for predicting the property with the molecular descriptors as predictive variables. The LFER can be used to construct models for predicting various specific properties from partition coefficients. Validations show that the models constructed according to the LFER have high predictive power and their performance is independent of the compounds used, including the models for the properties having little correlation with partition coefficients. The findings in this study are highly useful for applications in drug development and environmental safety.

Keywords: computational chemistry; linear free energy relationships; molecular properties; noncovalent interactions; partition coefficient; quantitative structure-property relationships.

FIGURE 1

Correlations between the molecular descriptor S_m and the effects of molecular size on various properties. (A) Linear associations between logP₁₆/logP_oct and S_m for alkanes. (B) Linear association between logK_brain (log of the partition coefficient from air to human brain) and S_m for nonpolar compounds. (C) Plot of the water to hexadecane phase-transferring free energy for depolarized solutes (ΔG_{tr_depol}) against the S_m values of the solutes.

FIGURE 2

Strong linear associations between the effects of HBAs on properties and HM_HBA. H_{M_HBA}: overall H-bonding capabilities of the HBAs of a solute. (A) For the property logP_oct (log of the partition coefficient between n-octanol and water). (B) For the property logP_chl (log of the partition coefficient between chloroform and water).

FIGURE 3

Prediction of organic solvent/water partition coefficients for validating the general LFER. (A, B) R² values of the simple regressions (gray columns) of logP₁₆(A)/logP_chl(B) against logP_oct and of the corresponding models constructed according to the LFER (black columns). HBA: compounds containing HBAs but no HBDs; HBD: compounds containing HBDs; apolar: nonpolar compounds; (C) Plot of observed logP₁₆ against the logP₁₆ calculated from the model constructed according to the LFER; (D) Plot of observed logP₁₆ against the logP₁₆ calculated from the model with logP_oct as predictive valuable.

FIGURE 4

Predictive power of models constructed according to the LFER. (A) External validation. Plots of observed logK_p (log of human skin permeability) against the logK_p calculated from the model constructed according to the LFER (logP_oct is used). (B) LOO cross-validation. Plots of observed logK_brain against the logK_brain calculated from the model constructed according to the LFER.