Intrinsic Aqueous Solubility: Mechanistically Transparent Data-Driven Modeling of Drug Substances.

Intrinsic aqueous solubility is a foundational property for understanding the chemical, technological, pharmaceutical, and environmental behavior of drug substances. Despite years of solubility research, molecular structure-based prediction of the intrinsic aqueous solubility of drug substances is still under active investigation. This paper describes the authors' systematic data-driven modelling in which two fit-for-purpose training data sets for intrinsic aqueous solubility were collected and curated, and three quantitative structure-property relationships were derived to make predictions for the most recent solubility challenge. All three models perform well individually, while being mechanistically transparent and easy to understand. Molecular descriptors involved in the models are related to the following key steps in the solubility process: dissociation of the molecule from the crystal, formation of a cavity in the solvent, and insertion of the molecule into the solvent. A consensus modeling approach with these models remarkably improved prediction capability and reduced the number of strong outliers by more than two times. The performance and outliers of the second solubility challenge predictions were analyzed retrospectively. All developed models have been published in the QsarDB.org repository according to FAIR principles and can be used without restrictions for exploring, downloading, and making predictions.

A role of flavonoids in cytochrome c-cardiolipin interactions.

The processes preceding the detachment of cytochrome c (cyt c) from the inner mitochondrial membrane in intrinsic apoptosis involve peroxidation of cardiolipin (CL) catalyzed by cyt c-CL complex. In the present work, we studied the effect of 17 dietary flavonoids on the peroxidase activity of cyt c bound to liposomes. Specifically, we explored the relationship between peroxidase activity and flavonoids' (1) potential to modulate cyt c unfolding, (2) effect on the oxidation state of heme iron, (3) membrane permeability, (4) membrane binding energy, and (5) structure. The measurements revealed that flavones, flavonols, and flavanols were the strongest, while isoflavones were the weakest inhibitors of the oxidation. Flavonoids' peroxidase inhibition activity correlated positively with their potential to suppress Trp-59 fluorescence in cyt c as well as the number of OH groups. Hydrophilic flavonoids, such as catechin, having the lowest membrane permeability and the strongest binding with phosphocholine (PC) based on the quantum chemical calculations exhibited the strongest inhibition of Amplex Red (AR) peroxidation, suggesting a membrane-protective function of flavonoids at the surface. The results of the present research specify basic principles for the design of molecules that will control the catalytic oxidation of lipids in mitochondrial membranes. These principles take into account the number of hydroxyl groups and hydrophilicity of flavonoids.

Logistic Classification Models for pH-Permeability Profile: Predicting Permeability Classes for the Biopharmaceutical Classification System.

Permeability is used to describe and evaluate the absorption of drug substances in the human gastrointestinal tract (GIT). Permeability is largely dependent on fluctuating pH that causes the ionization of drug substances and also influences regional absorption in the GIT. Therefore, classification models that characterize permeability at wide ranges of pH were derived in the current study. For this, drug substances were described with six data series that were measured with a parallel artificial membrane permeability assay (PAMPA), including a permeability profile at four pH values (3, 5, 7.4, and 9), and the highest and intrinsic membrane permeability. Logistic regression classification models were developed and compared by using two distinct sets of descriptors: (1) a hydrophobicity descriptor, the logarithm of the octanol-water partition (logPow) or distribution (logD) coefficient and (2) theoretical molecular descriptors. In both cases, models have good classification and descriptive capabilities for the training set (accuracy: 0.76-0.91). Triple validation with three sets of drug substances shows good prediction capability for all models: validation set (accuracy: 0.73-0.91), external validation set (accuracy: 0.72-0.9), and the permeability classes of FDA reference drugs for the biopharmaceutical classification system (BCS) (accuracy: 0.72-0.88). The identification of BCS permeability classes was further improved with decision trees that consolidated predictions from models with each descriptor type. These decision trees have higher confidence and accuracy (0.91 for theoretical molecular descriptors and 0.81 for hydrophobicity descriptors) than the individual models in assigning drug substances into BCS permeability classes. A detailed analysis of classification models and related decision trees suggests that they are suitable for predicting classes of permeability for passively transported drug substances, including specifically within the BCS framework. All developed models are available at the QsarDB repository ( http://dx.doi.org/10.15152/QDB.206 ).

pH-permeability profiles for drug substances: Experimental detection, comparison with human intestinal absorption and modelling.

The influence of pH on human intestinal absorption is frequently not considered in early drug discovery studies in the modelling and subsequent prediction of intestinal absorption for drug candidates. To bridge this gap, in this study, experimental membrane permeability data were measured for current and former drug substances with a parallel artificial membrane permeability assay (PAMPA) at different pH values (3, 5, 7.4 and 9). The presented data are in good agreement with human intestinal absorption, showing a clear influence of pH on the efficiency of intestinal absorption. For the measured data, simple and general quantitative structure-activity relationships (QSARs) were developed for each pH that makes it possible to predict the pH profiles for passive membrane permeability (i.e., a pH-permeability profile), and these predictions coincide well with the experimental data. QSARs are also proposed for the data series of highest and intrinsic membrane permeability. The molecular descriptors in the models were analysed and mechanistically related to the interaction pattern of permeability in membranes. In addition to the regression models, classification models are also proposed. All models were successfully validated and blind tested with external data. The models are available in the QsarDB repository (http://dx.doi.org/10.15152/QDB.203).

Quantitative structure-permeability relationships at various pH values for neutral and amphoteric drugs and drug-like compounds.

Human intestinal absorption is a key property for orally administered drugs and is dependent on pH. This study focuses on neutral and amphoteric compounds and their membrane permeabilities across the range of pH values found in the human intestine. The membrane permeability values for 15 neutral and 60 amphoteric compounds at pH 3, 5, 7.4 and 9 were measured using the parallel artificial membrane permeability assay (PAMPA). For each data series the quantitative structure-permeability relationships were developed and analysed. The results show that the membrane permeability of neutral compounds is attributed to a single structural characteristic, the hydrogen bond donor ability. Amphoteric compounds are more complex because of their chemical constitution, and therefore require three-parameter models to describe and predict membrane permeability. Analysis of the models for amphoteric compounds reveals that membrane permeability depends on multiple structural characteristics: the partition coefficient, hydrogen bond properties and the shape of the molecules. In addition to conventional validation strategies, two external compounds (isradipine and omeprazole) were tested and revealed very good agreement of pH profiles between experimental and predicted membrane permeability for all of the developed models. Selected QSAR models are available at the QsarDB repository (http://dx.doi.org/10.15152/QDB.184).

The Permeability of an Artificial Membrane for Wide Range of pH in Human Gastrointestinal Tract: Experimental Measurements and Quantitative StructureActivity Relationship.

In silico models for membrane permeability have been based on values measured for single pH. Depending on the diet (fasted/fed state) and part of human intestine the range of pH varies approximately from 2.4 to 8.0. This motivated to study and model the membrane permeability of chemicals considering the whole range of pH in the human intestine. For this, effective membrane permeability values were measured for 65 drugs and drug-like compounds using PAMPA method at four pHs (3, 5, 7.4, 9) over 48 h, introducing technological innovations for the time-dependence measurement. The highest permeability value of a compound from four pHs was used to derive QSAR analyzing a large pool of molecular descriptors and introducing new descriptor. Using stepwise forward selection approach a significant QSAR model was derived that included only two mechanistically relevant descriptors, the logarithmic octanol-water partition coefficient and hydrogen bonding surface area. Prediction confidence of the model was blind tested with a true external validation set of 15 compounds. The resulting QSAR model shows potential to combine permeability values from various pH-s into one descriptive and predictive model for estimating maximum permeability in human gastrointestinal tract. The QSAR model and data are available through the QsarDB repository (http://dx.doi.org/10.15152/QDB.137).

DrugLogit: logistic discrimination between drugs and nondrugs including disease-specificity by assigning probabilities based on molecular properties.

The increasing knowledge of both structure and activity of compounds provides a good basis for enhancing the pharmacological characterization of chemical libraries. In addition, pharmacology can be seen as incorporating both advances from molecular biology as well as chemical sciences, with innovative insight provided from studying target-ligand data from a ligand molecular point of view. Predictions and profiling of libraries of drug candidates have previously focused mainly on certain cases of oral bioavailability. Inclusion of other administration routes and disease-specificity would improve the precision of drug profiling. In this work, recent data are extended, and a probability-based approach is introduced for quantitative and gradual classification of compounds into categories of drugs/nondrugs, as well as for disease- or organ-specificity. Using experimental data of over 1067 compounds and multivariate logistic regressions, the classification shows good performance in training and independent test cases. The regressions have high statistical significance in terms of the robustness of coefficients and 95% confidence intervals provided by a 1000-fold bootstrapping resampling. Besides their good predictive power, the classification functions remain chemically interpretable, containing only one to five variables in total, and the physicochemical terms involved can be easily calculated. The present approach is useful for an improved description and filtering of compound libraries. It can also be applied sequentially or in combinations of filters, as well as adapted to particular use cases. The scores and equations may be able to suggest possible routes for compound or library modification. The data is made available for reuse by others, and the equations are freely accessible at http://hermes.chem.ut.ee/~alfx/druglogit.html.

Disease-Specific Differentiation Between Drugs and Non-Drugs Using Principal Component Analysis of Their Molecular Descriptor Space.

The physicochemical descriptor space has been extensively mapped and described in the literature for orally administered drugs and lead compounds. However, consideration of negative examples (non-drugs) or disease pathophysiology is not common in many studies. In the present work, a principal component analysis was carried out using drugs and non-drugs taking into account disease- and organ-specific categories, as well as different administration routes in addition to oral. The study involves 1386 relevant small-molecules including natural and synthetic products. Drug-specific as well as disease-category-specific or organ-specific regions and their respective threshold sets (ranges of descriptors) relative to non-drugs were elucidated on the scores plot and validated with external, independent sets of drugs and non-drugs. The respective loadings plot of molecular descriptors was rationalized in terms of physicochemically relevant groups related to the components of solvation free energy. The results of this analysis can contribute to the improved profiling of drug candidates and libraries making use of disease- and organ-specificity coded by physicochemical descriptors and ligand binding efficiency.