Octanol/water partition coefficients estimated using retention times in reverse-phase liquid chromatography and calculated in silico as one of the determinant factors for pharmacokinetic parameter estimations of general chemical substances.

The octanol/water partition coefficient P (logP) is a hydrophobicity index and is one of the determining factors for the pharmacokinetics of orally administered substances because it influences membrane permeability. To illustrate the wide-ranging variety of compounds in the chemical space, a two-dimensional data plot consisting of 25 blocks was previously proposed based on a substance's in silico chemical descriptors. The logP values of approximately 200 diverse chemicals (test plus reference compounds covering all 25 blocks of the chemical space) were estimated experimentally using retention times in reverse-phase liquid chromatography; these values were compared with those of authentic reference compounds with established logP values (available for 17 of 60 reference substances in the Organization for Economic Co-operation and Development Test Guideline 117). The logP values of 140 of 165 chemicals successfully estimated using four different mobile phase conditions (pH 2, 4, 7, and 10 for molecular forms) correlated significantly with those calculated using the in silico packages ChemDraw and ACD/Percepta (r > 0.72). Although substances that neighbored authentic compounds in the chemical space had precisely correlated logP values estimated experimentally and in silico, some compounds that were more distant from authentic substances showed lower logP values than those estimated in silico. These results indicate that additional authentic reference materials with wider ranging chemical diversity and their logP values from reverse-phase liquid chromatography should be included in the international test guidance to promote simple and reliable estimation of octanol/water partition coefficients, which are important determinant factors for the pharmacokinetics of general chemicals.

Screening and Validation of Odorants against Influenza A Virus Using Interpretable Regression Models.

Influenza is a respiratory infection caused by the influenza virus that is prevalent worldwide. One of the most contagious variants of influenza is influenza A virus (IAV), which usually spreads in closed spaces through aerosols. Preventive measures such as novel compounds are needed that can act on viral membranes and provide a safe environment against IAV infection. In this study, we screened compounds with common fragrances that are generally used to mask unpleasant odors but can also exhibit antiviral activity against a strain of IAV. Initially, a set of 188 structurally diverse odorants were collected, and their antiviral activity was measured in vapor phase against the IAV solution. Regression models were built for the prediction of antiviral activity using this set of odorants by taking into account their structural features along with vapor pressure and partition coefficient (n-octanol/water). The models were interpreted using a feature weighting approach and Shapley Additive exPlanations to rationalize the predictions as an additional validation for virtual screening. This model was used to screen odorants from an in-house odorant data set consisting of 2020 odorants, which were later evaluated using in vitro experiments. Out of 11 odorants proposed using the final model, 8 odorants were found to exhibit antiviral activity. The feature interpretation of screened odorants suggested that they contained hydrophilic substructures, such as hydroxyl group, which might contribute to denaturation of proteins on the surface of the virus. These odorants should be explored as a preventive measure in closed spaces to decrease the risk of infections of IAV.

Technique of Augmenting Molecular Graph Data by Perturbating Hidden Features.

Quantitative structure-property relationship models are useful in efficiently searching for molecules with desired properties in drug discovery and materials development. In recent years, many such models based on graph neural networks, showing good prediction performance, have been reported. Training graph neural networks generally require many samples, but by using a training method for a small dataset, it is possible to extract features that enable successful prediction. Herein, we design a method of augmenting graph data. In this method, random perturbations are added with a certain probability to some vertex features during message passing. We verify the proposed method's effectiveness in regression and classification tasks. It is confirmed that the proposed method is effective when the perturbation is added immediately before the readout of the graph neural network, and the effect of the data augmentation is most evident for small datasets of approximately 1000 samples.

Prediction of Reaction Yield for Buchwald-Hartwig Cross-coupling Reactions Using Deep Learning.

Chemical reaction yield is one of the most important factors for determining reaction conditions. Recently, several machine learning-based prediction models using high-throughput experiment (HTE) data sets were reported for the prediction of reaction yield. However, none of them were at a practical level in terms of predictive ability. In this study, we propose a message passing neural network (MPNN) model for chemical yield prediction, focusing on the Buchwald-Hartwig cross-coupling HTE data set. As an initial atom embedding in MPNN model, we propose to use the Mol2Vec feature vectors pre-trained using a large compound database. Predictive ability of the proposed model was higher than that of previously reported five models for the three out of five data sets. Moreover, visualization of important atoms based on self-attention mechanism was in favor of Mol2Vec as an atom embedding rather than other embeddings including previously employed simple representations.

An Updated In Silico Prediction Method for Volumes of Systemic Circulation of 323 Disparate Chemicals for Use in Physiologically Based Pharmacokinetic Models to Estimate Plasma and Tissue Concentrations after Oral Doses in Rats.

Updated algorithms for predicting the volumes of systemic circulation (V1), along with absorption rate constants and hepatic intrinsic clearances, as input parameters for physiologically based pharmacokinetic (PBPK) models were established to improve the accuracy of estimated plasma and tissue concentrations of 323 chemicals after virtual oral administrations in rats. Using ridge regression with an enlarged set of chemical descriptors (up to 99), the estimated input V1 values resulted in an improved correlation coefficient (from 246 compounds) with the traditionally determined values. The PBPK model input parameters for rats of diverse compounds can be precisely estimated by increasing the number of descriptors.

Prediction of permeability across intestinal cell monolayers for 219 disparate chemicals using in vitro experimental coefficients in a pH gradient system and in silico analyses by trivariate linear regressions and machine learning.

For medicines, the apparent membrane permeability coefficients (Papp) across human colorectal carcinoma cell line (Caco-2) monolayers under a pH gradient generally correlate with the fraction absorbed after oral intake. Furthermore, the in vitro Papp values of 29 industrial chemicals were found to have an inverse association with their reported no-observed effect levels for hepatotoxicity in rats. In the current study, we expanded our influx permeability predictions for the 90 previously investigated chemicals to both influx and efflux permeability predictions for 207 diverse primary compounds, along with those for 23 secondary compounds. Trivariate linear regression analysis found that the observed influx and efflux logPapp values determined by in vitro experiments significantly correlated with molecular weights and the octanol-water distribution coefficients at apical and basal pH levels (pH 6.0 and 7.4, respectively) (apical to basal, r = 0.76, n = 198; and basal to apical, r = 0.77, n = 202); the distribution coefficients were estimated in silico. Further, prediction accuracy was enhanced by applying a light gradient boosting machine learning system (LightGBM) to estimate influx and efflux logPapp values that incorporated 17 and 19 in silico chemical descriptors (r = 0.83-0.84, p < 0.001). The determination in vitro and/or prediction in silico of permeability coefficients across intestinal cell monolayers of a diverse range of industrial chemicals/food components/medicines could contribute to the safety evaluations of oral intakes of general chemicals in humans. Such new alternative methods could also reduce the need for animal testing during toxicity assessment.

Interpretation of Ligand-Based Activity Cliff Prediction Models Using the Matched Molecular Pair Kernel.

Activity cliffs (ACs) are formed by two structurally similar compounds with a large difference in potency. Accurate AC prediction is expected to help researchers' decisions in the early stages of drug discovery. Previously, predictive models based on matched molecular pair (MMP) cliffs have been proposed. However, the proposed methods face a challenge of interpretability due to the black-box character of the predictive models. In this study, we developed interpretable MMP fingerprints and modified a model-specific interpretation approach for models based on a support vector machine (SVM) and MMP kernel. We compared important features highlighted by this SVM-based interpretation approach and the SHapley Additive exPlanations (SHAP) as a major model-independent approach. The model-specific approach could capture the difference between AC and non-AC, while SHAP assigned high weights to the features not present in the test instances. For specific MMPs, the feature weights mapped by the SVM-based interpretation method were in agreement with the previously confirmed binding knowledge from X-ray co-crystal structures, indicating that this method is able to interpret the AC prediction model in a chemically intuitive manner.

Sparse Topological Pharmacophore Graphs for Interpretable Scaffold Hopping.

The aim of scaffold hopping (SH) is to find compounds consisting of different scaffolds from those in already known active compounds, giving an opportunity for unexplored regions of chemical space. We previously demonstrated the usefulness of pharmacophore graphs (PhGs) for this purpose through proof-of-concept virtual screening experiments. PhGs consist of nodes and edges corresponding to pharmacophoric features (PFs) and their topological distances. Although PhGs were effective in SH, they are hard to interpret as they are complete graphs. Herein, we introduce an intuitive representation of a molecule, termed as sparse pharmacophore graphs (SPhG) by keeping the topological distances among PFs as much as possible while reducing the number of edges in the graphs. Several benchmark calculations quantitatively confirmed the sparseness of the graphs and the preservation of topological distances among pharmacophoric points. As proof-of-concept applications, virtual screening (VS) trials for SH were conducted using active and inactive compounds from ChEMBL and PubChem databases for three biological targets: thrombin, tyrosine kinase ABL1, and κ-opioid receptor. The performances of VS were comparable with using fully connected PhGs. Furthermore, highly ranked SPhGs were interpretable for the three biological targets, in particular for thrombin, for which selected SPhGs were in agreement with the structure-based interpretation.

Ranking-Oriented Quantitative Structure-Activity Relationship Modeling Combined with Assay-Wise Data Integration.

In ligand-based drug design, quantitative structure-activity relationship (QSAR) models play an important role in activity prediction. One of the major end points of QSAR models is half-maximal inhibitory concentration (IC50). Experimental IC50 data from various research groups have been accumulated in publicly accessible databases, providing an opportunity for us to use such data in predictive QSAR models. In this study, we focused on using a ranking-oriented QSAR model as a predictive model because relative potency strength within the same assay is solid information that is not based on any mechanical assumptions. We conducted rigorous validation using the ChEMBL database and previously reported data sets. Ranking support vector machine (ranking-SVM) models trained on compounds from similar assays were as good as support vector regression (SVR) with the Tanimoto kernel trained on compounds from all the assays. As effective ways of data integration, for ranking-SVM, integrated compounds should be selected from only similar assays in terms of compounds. For SVR with the Tanimoto kernel, entire compounds from different assays can be incorporated.

In Silico Prediction of Input Parameters for Simplified Physiologically Based Pharmacokinetic Models for Estimating Plasma, Liver, and Kidney Exposures in Rats after Oral Doses of 246 Disparate Chemicals.

Recently developed computational models can estimate plasma, hepatic, and renal concentrations of industrial chemicals in rats. Typically, the input parameter values (i.e., the absorption rate constant, volume of systemic circulation, and hepatic intrinsic clearance) for simplified physiologically based pharmacokinetic (PBPK) model systems are calculated to give the best fit to measured or reported in vivo blood substance concentration values in animals. The purpose of the present study was to estimate in silico these three input pharmacokinetic parameters using a machine learning algorithm applied to a broad range of chemical properties obtained from several cheminformatics software tools. These in silico estimated parameters were then incorporated into PBPK models for predicting internal exposures in rats. Following this approach, simplified PBPK models were set up for 246 drugs, food components, and industrial chemicals with a broad range of chemical structures. We had previously generated PBPK models for 158 of these substances, whereas 88 for which concentration series data were available in the literature were newly modeled. The values for the absorption rate constant, volume of systemic circulation, and hepatic intrinsic clearance could be generated in silico by equations containing between 14 and 26 physicochemical properties. After virtual oral dosing, the output concentration values of the 246 compounds in plasma, liver, and kidney from rat PBPK models using traditionally determined and in silico estimated input parameters were well correlated (r ≥ 0.83). In summary, by using PBPK models consisting of chemical receptor (gut), metabolizing (liver), excreting (kidney), and central (main) compartments with in silico-derived input parameters, the forward dosimetry of new chemicals could provide the plasma/tissue concentrations of drugs and chemicals after oral dosing, thereby facilitating estimates of hematotoxic, hepatotoxic, or nephrotoxic potential as a part of risk assessment.

Investigation of Preprocessing and Validation Methodologies for PAT: Case Study of the Granulation and Coating Steps for the Manufacturing of Ethenzamide Tablets.

After the Food and Drug Association in the USA published guidelines on the enhanced use of process analytical technology (PAT) and continuous manufacturing, many studies regarding PAT and continuous manufacturing have been published. This paper describes a case study involving granulation and coating steps with ethenzamide to investigate interference for PAT model construction and model management. We investigated what factors should be considered and addressed when PAT is implemented for continuous manufacturing and how predictive models should be constructed. The product qualities that were monitored were moisture content and particle size in the granulation step and tablet weight and moisture content in the coating step. We have constructed models for the granulation step and validated the predictive capability of the models against an external dataset. A partial least squares (PLS) model with manual wavelength selection had the best predictive accuracy for loss on drying against the external validation set. We found that the prediction of loss on drying was accurate, but the prediction of particle size was not sufficiently accurate. In the coating step, because of the small amount of data, we performed three-fold cross-validation and y-scrambling 10 times, to select the optimal hyper-parameters and to check if the models were fitted to chance correlations. We confirmed that the coating agent weights, tablet weights, and water content could be accurately predicted based on the mean of the R2 score for cross-validation. Addition of other variables, as well as the absorbance, slightly improved the predictive accuracy.

Comparing predictive ability of QSAR/QSPR models using 2D and 3D molecular representations.

Quantitative structure-activity relationship (QSAR) and quantitative structure-property relationship (QSPR) models predict biological activity and molecular property based on the numerical relationship between chemical structures and activity (property) values. Molecular representations are of importance in QSAR/QSPR analysis. Topological information of molecular structures is usually utilized (2D representations) for this purpose. However, conformational information seems important because molecules are in the three-dimensional space. As a three-dimensional molecular representation applicable to diverse compounds, similarity between a test molecule and a set of reference molecules has been previously proposed. This 3D representation was found to be effective on virtual screening for early enrichment of active compounds. In this study, we introduced the 3D representation into QSAR/QSPR modeling (regression tasks). Furthermore, we investigated relative merits of 3D representations over 2D in terms of the diversity of training data sets. For the prediction task of quantum mechanics-based properties, the 3D representations were superior to 2D. For predicting activity of small molecules against specific biological targets, no consistent trend was observed in the difference of performance using the two types of representations, irrespective of the diversity of training data sets.

Improvement of the Structure Generator DAECS with Respect to Structural Diversity.

The development of novel organic compounds with desired properties is time consuming and costly. Thus, the quantitative structure-property relationship (QSPR) model is used widely for efficiently discovering compounds with the desired properties. Novel structures can be generated from a variety of input structures in silico by structure generators. We previously developed the structure generator DAECS to yield highly active drug-like structures. However, the structural diversity of the structures generated by DAECS was still small for practical applications such as drug discovery. In this paper, we present structure modification rules and the algorithm to output more diverse structures through the DAECS workflow. Two new types of structural modification rules, bond contraction and ring mergence, were added. The new algorithm, which restricts the search area and subsequently clusters structures on a two-dimensional map generated by generative topographic mapping, was implemented for the repetitive selection of seed structures. A case study was conducted to evaluate our method using ligand structures for the histamine H1 receptor. The results showed improved structural diversity than the previous method.

Ligand-based Activity Cliff Prediction Models with Applicability Domain.

Activity cliffs (ACs) are formed by pairs of structurally similar compounds with large differences in potency. Predicting ACs is of high interest in lead optimization for drug discovery. Previous AC prediction models that focused on matched molecular pair (MMP) cliffs produced adequate performances. However, the extrapolation ability of these models is unclear because the main scaffold for MMPs, the core structure, could exist in both training and test data sets. Also, representation of MMPs did not consider the attachment points where the core and R-group substituents are connected. In this study, we aimed to improve a ligand-based AC prediction method using molecular fingerprints. We incorporated applicability domain, which was defined using R-path fingerprints to consider the local environment around an attachment point. Rigorous evaluation of the extrapolation ability of AC prediction models showed that MMP-cliffs were accurately predicted for nine biological targets. Furthermore, incorporation of training MMPs with cores distinct from those of test MMPs improved the predictability compared with using training MMPs with only similar cores.

Physiologically Based Pharmacokinetic Models Predicting Renal and Hepatic Concentrations of Industrial Chemicals after Virtual Oral Doses in Rats.

Recently developed high-throughput in vitro assays in combination with computational models could provide alternatives to animal testing. The purpose of the present study was to model the plasma, hepatic, and renal pharmacokinetics of approximately 150 structurally varied types of drugs, food components, and industrial chemicals after virtual external oral dosing in rats and to determine the relationship between the simulated internal concentrations in tissue/plasma and their lowest-observed-effect levels. The model parameters were based on rat plasma data from the literature and empirically determined pharmacokinetics measured after oral administrations to rats carried out to evaluate hepatotoxic or nephrotic potentials. To ensure that the analyzed substances exhibited a broad diversity of chemical structures, their structure-based location in the chemical space underwent projection onto a two-dimensional plane, as reported previously, using generative topographic mapping. A high-throughput in silico one-compartment model and a physiologically based pharmacokinetic (PBPK) model consisting of chemical receptor (gut), metabolizing (liver), central (main), and excreting (kidney) compartments were developed in parallel. For 159 disparate chemicals, the maximum plasma concentrations and the areas under the concentration-time curves obtained by one-compartment models and modified simple PBPK models were closely correlated. However, there were differences between the PBPK modeled and empirically obtained hepatic/renal concentrations and plasma maximal concentrations/areas under the concentration-time curves of the 159 chemicals. For a few compounds, the lowest-observed-effect levels were available for hepatotoxicity and nephrotoxicity in the Hazard Evaluation Support System Integrated Platform in Japan. The areas under the renal or hepatic concentration-time curves estimated using PBPK modeling were inversely associated with these lowest-observed-effect levels. Using PBPK forward dosimetry could provide the plasma/tissue concentrations of drugs and chemicals after oral dosing, thereby facilitating estimates of nephrotoxic or hepatotoxic potential as a part of the risk assessment.

Exploring Topological Pharmacophore Graphs for Scaffold Hopping.

The primary goal of ligand-based virtual screening is to identify active compounds consisting of a core scaffold that is not found in the current active compound pool. Scaffold hopping is the term used for this purpose. In the present study, topological representations of pharmacophore features on chemical graphs were investigated for scaffold hopping. Pharmacophore graphs (PhGs), which consist of pharmacophore features as nodes and their topological distances as edges, were used as a representation of important information on compounds being active. We investigated ranking methods for prioritizing PhGs for scaffold hopping. The proposed method, NScaffold, which ranks PhGs based on the number of scaffolds covered by the PhGs, outperforms other conventional methods. As a demonstrative case, using a thrombin inhibitor data set, we interpreted the highest-ranked PhGs by NScaffold from the protein-ligand interaction point of view. It resulted that the NScaffold method successfully retrieved three known important interactions, showing the potential for identifying scaffold-hopped compounds with interpretable PhGs.

Application of the mol2vec Technology to Large-size Data Visualization and Analysis.

Generative Topographic Mapping (GTM) is a dimensionality reduction method, which is widely used for both data visualization and structure-activity modeling. Large dimensionality of the initial data space may require significant computational resources and slow down the GTM construction. Therefore, it may be meaningful to reduce the number of descriptors used for encoding molecular structures. The Principal Component Analysis (PCA), a standard preprocessing tool, suffers from the information loss upon the dimensionality reduction. As an alternative, we propose to use substructure vector embedding provided by the mol2vec technique. In addition to the data dimensionality reduction, this technology also accounts for proximity of substructures in molecular graphs. In this study, dimensionality of large descriptor spaces of ISIDA fragment descriptors or Morgan fingerprints were reduced using either the PCA or the mol2vec method. The latter significantly speeds up GTM training without compromising its predictive power in bioactivity classification tasks.

Determination and prediction of permeability across intestinal epithelial cell monolayer of a diverse range of industrial chemicals/drugs for estimation of oral absorption as a putative marker of hepatotoxicity.

Apparent permeability coefficients (P app) across a human intestinal epithelial Caco-2 cell monolayer were measured for a range of industrial/drug chemicals. A predictive equation for determining in vitro P app values of fifty-six substances was set up using multivariate regression analysis based on in silico-estimated physicochemical properties (molecular weights and water distribution coefficients for apical and basal pH environments) (r = 0.77, p <  0.01). Predicted logP app values of a secondary set of 34 compounds were correlated with the measured values. Under the medicinal logP app values associated with their reported fraction absorbed, a significant inverse non-linear correlation was found between the logarithmic transformed values of observed P app values and reported hepatic no-observed-effect levels of industrial chemicals (r = -0.55, p <  0.01, n = 29). In vitro determination and/or in silico prediction of permeability across intestinal cells could be effective for estimating oral absorption as a putative indicator for hepatotoxicity.

Comparison and improvement of the predictability and interpretability with ensemble learning models in QSPR applications.

Ensemble learning helps improve machine learning results by combining several models and allows the production of better predictive performance compared to a single model. It also benefits and accelerates the researches in quantitative structure-activity relationship (QSAR) and quantitative structure-property relationship (QSPR). With the growing number of ensemble learning models such as random forest, the effectiveness of QSAR/QSPR will be limited by the machine's inability to interpret the predictions to researchers. In fact, many implementations of ensemble learning models are able to quantify the overall magnitude of each feature. For example, feature importance allows us to assess the relative importance of features and to interpret the predictions. However, different ensemble learning methods or implementations may lead to different feature selections for interpretation. In this paper, we compared the predictability and interpretability of four typical well-established ensemble learning models (Random forest, extreme randomized trees, adaptive boosting and gradient boosting) for regression and binary classification modeling tasks. Then, the blending methods were built by summarizing four different ensemble learning methods. The blending method led to better performance and a unification interpretation by summarizing individual predictions from different learning models. The important features of two case studies which gave us some valuable information to compound properties were discussed in detail in this report. QSPR modeling with interpretable machine learning techniques can move the chemical design forward to work more efficiently, confirm hypothesis and establish knowledge for better results.