Exploring the octanol-water partition coefficient dataset using deep learning techniques and data augmentation

Temperature dependence of the rain-gas and snow-gas partition coefficients for nearly a thousand chemicals

Compared to the particle-gas partition coefficients (KPG), the rain-gas (KRG) and snow-gas (KSG) partition coefficients are also essential in studying the environmental behavior and fate of chemicals in the atmosphere. While the temperature dependence for the KPG have been extensively studied, the study for KRG and KSG are still lacking. Adsorption coefficients between water surface-air (KIA) and snow surface-air (KJA), as well as partition coefficients between water-air (KWA) and octanol-air (KOA) are vital in calculating KRG and KSG. These four basic adsorption and partition coefficients are also temperature-dependent, given by the well-known two-parameters Antoine equation logKXY = AXY + BXY/T, where KXY is the adsorption or partition coefficients, AXY and BXY are Antoine parameters (XY stand for IA, JA, WA, and OA), and T is the temperature in Kelvin. In this study, the parameters AXY and BXY are calculated for 943 chemicals, and logKXY can be estimated at any ambient temperature for these chemicals using these Antoine parameters. The results are evaluated by comparing these data with published experimental and modeled data, and the results show reasonable accuracy. Based on these coefficients, temperature-dependence of logKRG and logKSG is studied. It is found that both logKRG and logKSG are linearly related to 1/T, and Antoine parameters for logKRG and logKSG are also estimated. Distributions of the 943 chemicals in the atmospheric phases (gas, particle, and rain/snow), are illustrated in a Chemical Space Map. The findings reveal that, at environmental temperatures and precipitation days, the dominant state for the majority of chemicals is the gaseous phase. All the AXY and BXY values for logKSG, logKRG, and basic adsorption and partition coefficients, both modeled by this study and collected from published work, are systematically organized into an accessible dataset for public utilization.

Can membrane permeability of zwitterionic compounds be predicted by the solubility-diffusion model?

Zwitterions contain both positively and negatively charged functional groups, resulting in an overall net neutral charge. Nevertheless, the membrane permeability of the zwitterionic form of a compound is assumed to be much lower than the permeability of the uncharged neutral form. Although a significant proportion of pharmaceuticals are zwitterionic, it has not been clear so far whether their permeability is dominated by the permeation of the zwitterionic or the neutral form, since neutral fractions are often quite low as compared to the zwitterionic fraction. This complicates the in silico prediction of the permeability of zwitterionic compounds. In this work, we re-evaluated existing in vitro permeability data from literature measured with Caco-2/MDCK cell assays, using more strict exclusion criteria for effects like diffusion limitation by the aqueous boundary layers, paracellular transport, active transport and retention. Using this re-evaluated data set, we show that extracted intrinsic permeabilities of the neutral fraction are well predicted by the solubility-diffusion model (RMSE = 1.21; n = 18) if the permeability of the zwitterionic species is assumed negligible. Our work thus suggests that only the neutral species is relevant for the membrane permeability of zwitterionic compounds, and that membrane permeability of zwitterionic compounds is indeed predictable by the solubility-diffusion model.

Machine learning framework for predicting cytotoxicity and identifying toxicity drivers of disinfection byproducts

Drinking water disinfection can result in the formation disinfection byproducts (DBPs, > 700 have been identified to date), many of them are reportedly cytotoxic, genotoxic, or developmentally toxic. Analyzing the toxicity levels of these contaminants experimentally is challenging, however, a predictive model could rapidly and effectively assess their toxicity. In this study, machine learning models were developed to predict DBP cytotoxicity based on their chemical information and exposure experiments. The Random Forest model achieved the best performance (coefficient of determination of 0.62 and root mean square error of 0.63) among all the algorithms screened. Also, the results of a probabilistic model demonstrated reliable model predictions. According to the model interpretation, halogen atoms are the most prominent features for DBP cytotoxicity compared to other chemical substructures. The presence of iodine and bromine is associated with increased cytotoxicity levels, while the presence of chlorine is linked to a reduction in cytotoxicity levels. Other factors including chemical substructures (CC, N, CN, and 6-member ring), cell line, and exposure duration can significantly affect the cytotoxicity of DBPs. The similarity calculation indicated that the model has a large applicability domain and can provide reliable predictions for DBPs with unknown cytotoxicity. Finally, this study showed the effectiveness of data augmentation in the scenario of data scarcity.

Pitfalls in evaluating permeability experiments with Caco-2/MDCK cell monolayers

When studying the transport of molecules across biological membranes, intrinsic membrane permeability (P0) is more informative than apparent permeability (Papp), because it eliminates external (setup-specific) factors, provides consistency across experiments and mechanistic insight. It is thus an important building block for modeling the total permeability in any given scenario. However, extracting P0 is often difficult, if not impossible, when the membrane is not the dominant transport resistance. In this work, we set out to analyze Papp values measured with Caco-2/MDCK cell monolayers of 69 literature references. We checked the Papp values for a total of 318 different compounds for the extractability of P0, considering possible limitations by aqueous boundary layers, paracellular transport, recovery issues, active transport, a possible proton flux limitation, and sink conditions. Overall, we were able to extract 77 reliable P0 values, which corresponds to about one quarter of the total compounds analyzed, while about half were limited by the diffusion through the aqueous layers. Compared to an existing data set of P0 values published by Avdeef, our approach resulted in a much higher exclusion of compounds. This is a consequence of stricter compound- and reference-specific exclusion criteria, but also because we considered possible concentration-shift effects due to different pH values in the aqueous layers, an effect only recently described in literature. We thus provide a consistent and reliable set of P0, e.g. as a basis for future modeling.

Comment on Naef, R.; Acree, W.E., Jr. Calculation of the Three Partition Coefficients logPow, logKoa and logKaw of Organic Molecules at Standard Conditions at Once by Means of a Generally Applicable Group Additivity Method. Preprints, 2023120275

Next to the paper referred to in the title [...].

ConfSolv: Prediction of Solute Conformer-Free Energies across a Range of Solvents

Predicting Gibbs free energy of solution is key to understanding the solvent effects on thermodynamics and reaction rates for kinetic modeling. Accurately computing solution free energies requires the enumeration and evaluation of relevant solute conformers in solution. However, even after generation of relevant conformers, determining their free energy of solution requires an expensive workflow consisting of several ab initio computational chemistry calculations. To help address this challenge, we generate a large data set of solution free energies for nearly 44,000 solutes with almost 9 million conformers calculated in 41 different solvents using density functional theory and COSMO-RS and quantify the impact of solute conformers on the solution free energy. We then train a message passing neural network to predict the relative solution free energies of a set of solute conformers, enabling the identification of a small subset of thermodynamically relevant conformers. The model offers substantial computational time savings with predictions usually substantially within 1 kcal/mol of the free energy of the solution calculated by using computational chemical methods.

Recent Studies of Artificial Intelligence on In Silico Drug Absorption

Absorption is an important area of research in pharmacochemistry and drug development, because the drug has to be absorbed before any drug effects can occur. Furthermore, the ADMET (Absorption, Distribution, Metabolism, Excretion, and Toxicity) profile of drugs can be directly and considerably altered by modulating factors affecting absorption. Many drugs in development fail because of poor absorption. The research and continuous efforts of researchers in recent years have brought many successes and promises in drug absorption property prediction, especially in silico, which helps to reduce the time and cost significantly for screening undesirable drug candidates. In this report, we explicitly provide an overview of recent in silico studies on predicting absorption properties, especially from 2019 to the present, using artificial intelligence. Additionally, we have collected and investigated public databases that support absorption prediction research. On those grounds, we also proposed the challenges and development directions of absorption prediction in the future. We hope this review can provide researchers with valuable guidelines on absorption prediction to facilitate the development of newer approaches in drug discovery.

Open-Source Machine Learning in Computational Chemistry

The field of computational chemistry has seen a significant increase in the integration of machine learning concepts and algorithms. In this Perspective, we surveyed 179 open-source software projects, with corresponding peer-reviewed papers published within the last 5 years, to better understand the topics within the field being investigated by machine learning approaches. For each project, we provide a short description, the link to the code, the accompanying license type, and whether the training data and resulting models are made publicly available. Based on those deposited in GitHub repositories, the most popular employed Python libraries are identified. We hope that this survey will serve as a resource to learn about machine learning or specific architectures thereof by identifying accessible codes with accompanying papers on a topic basis. To this end, we also include computational chemistry open-source software for generating training data and fundamental Python libraries for machine learning. Based on our observations and considering the three pillars of collaborative machine learning work, open data, open source (code), and open models, we provide some suggestions to the community.

Development and test of highly accurate endpoint free energy methods. 2: Prediction of logarithm of n-octanol-water partition coefficient (logP) for druglike molecules using MM-PBSA method

The logarithm of n-octanol-water partition coefficient (logP) is frequently used as an indicator of lipophilicity in drug discovery, which has substantial impacts on the absorption, distribution, metabolism, excretion, and toxicity of a drug candidate. Considering that the experimental measurement of the property is costly and time-consuming, it is of great importance to develop reliable prediction models for logP. In this study, we developed a transfer free energy-based logP prediction model-FElogP. FElogP is based on the simple principle that logP is determined by the free energy change of transferring a molecule from water to n-octanol. The underlying physical method to calculate transfer free energy is the molecular mechanics-Poisson Boltzmann surface area (MM-PBSA), thus this method is named as free energy-based logP (FElogP). The superiority of FElogP model was validated by a large set of 707 structurally diverse molecules in the ZINC database for which the measurement was of high quality. Encouragingly, FElogP outperformed several commonly-used QSPR or machine learning-based logP models, as well as some continuum solvation model-based methods. The root-mean-square error (RMSE) and Pearson correlation coefficient (R) between the predicted and measured values are 0.91 log units and 0.71, respectively, while the runner-up, the logP model implemented in OpenBabel had an RMSE of 1.13 log units and R of 0.67. Given the fact that FElogP was not parameterized against experimental logP directly, its excellent performance is likely to be expanded to arbitrary organic molecules covered by the general AMBER force fields.

Using Machine Learning To Predict Partition Coefficient (Log P) and Distribution Coefficient (Log D) with Molecular Descriptors and Liquid Chromatography Retention Time

During preclinical evaluations of drug candidates, several physicochemical (p-chem) properties are measured and employed as metrics to estimate drug efficacy in vivo. Two such p-chem properties are the octanol-water partition coefficient, Log P, and distribution coefficient, Log D, which are useful in estimating the distribution of drugs within the body. Log P and Log D are traditionally measured using the shake-flask method and high-performance liquid chromatography. However, it is challenging to measure these properties for species that are very hydrophobic (or hydrophilic) owing to the very low equilibrium concentrations partitioned into octanol (or aqueous) phases. Moreover, the shake-flask method is relatively time-consuming and can require multistep dilutions as the range of analyte concentrations can differ by several orders of magnitude. Here, we circumvent these limitations by using machine learning (ML) to correlate Log P and Log D with liquid chromatography (LC) retention time (RT). Predictive models based on four ML algorithms, which used molecular descriptors and LC RTs as features, were extensively tested and compared. The inclusion of RT as an additional descriptor improves model performance (MAE = 0.366 and R² = 0.89), and Shapley additive explanations analysis indicates that RT has the highest impact on model accuracy.

Dimensionally reduced machine learning model for predicting single component octanol-water partition coefficients

MF-LOGP, a new method for determining a single component octanol-water partition coefficients ([Formula: see text]) is presented which uses molecular formula as the only input. Octanol-water partition coefficients are useful in many applications, ranging from environmental fate and drug delivery. Currently, partition coefficients are either experimentally measured or predicted as a function of structural fragments, topological descriptors, or thermodynamic properties known or calculated from precise molecular structures. The MF-LOGP method presented here differs from classical methods as it does not require any structural information and uses molecular formula as the sole model input. MF-LOGP is therefore useful for situations in which the structure is unknown or where the use of a low dimensional, easily automatable, and computationally inexpensive calculations is required. MF-LOGP is a random forest algorithm that is trained and tested on 15,377 data points, using 10 features derived from the molecular formula to make [Formula: see text] predictions. Using an independent validation set of 2713 data points, MF-LOGP was found to have an average [Formula: see text] = 0.77 ± 0.007, [Formula: see text] = 0.52 ± 0.003, and [Formula: see text] = 0.83 ± 0.003. This performance fell within the spectrum of performances reported in the published literature for conventional higher dimensional models ([Formula: see text] = 0.42-1.54, [Formula: see text] = 0.09-1.07, and [Formula: see text] = 0.32-0.95). Compared with existing models, MF-LOGP requires a maximum of ten features and no structural information, thereby providing a practical and yet predictive tool. The development of MF-LOGP provides the groundwork for development of more physical prediction models leveraging big data analytical methods or complex multicomponent mixtures.

Machine Learning for Fast, Quantum Mechanics-Based Approximation of Drug Lipophilicity

Lipophilicity, as measured by the partition coefficient between octanol and water (log P), is a key parameter in early drug discovery research. However, measuring log P experimentally is difficult for specific compounds and log P ranges. The resulting lack of reliable experimental data impedes development of accurate in silico models for such compounds. In certain discovery projects at Novartis focused on such compounds, a quantum mechanics (QM)-based tool for log P estimation has emerged as a valuable supplement to experimental measurements and as a preferred alternative to existing empirical models. However, this QM-based approach incurs a substantial computational cost, limiting its applicability to small series and prohibiting quick, interactive ideation. This work explores a set of machine learning models (Random Forest, Lasso, XGBoost, Chemprop, and Chemprop3D) to learn calculated log P values on both a public data set and an in-house data set to obtain a computationally affordable, QM-based estimation of drug lipophilicity. The message-passing neural network model Chemprop emerged as the best performing model with mean absolute errors of 0.44 and 0.34 log units for scaffold split test sets of the public and in-house data sets, respectively. Analysis of learning curves suggests that a further decrease in the test set error can be achieved by increasing the training set size. While models directly trained on experimental data perform better at approximating experimentally determined log P values than models trained on calculated values, we discuss the potential advantages of using calculated log P values going beyond the limits of experimental quantitation. We analyze the impact of the data set splitting strategy and gain insights into model failure modes. Potential use cases for the presented models include pre-screening of large compound collections and prioritization of compounds for full QM calculations.

Multitask Deep Ensemble Prediction of Molecular Energetics in Solution: From Quantum Mechanics to Experimental Properties

The past few years have witnessed significant advances in developing machine learning methods for molecular energetics predictions, including calculated electronic energies with high-level quantum mechanical methods and experimental properties, such as solvation free energy and logP. Typically, task-specific machine learning models are developed for distinct prediction tasks. In this work, we present a multitask deep ensemble model, sPhysNet-MT-ens5, which can simultaneously and accurately predict electronic energies of molecules in gas, water, and octanol phases, as well as transfer free energies at both calculated and experimental levels. On the calculated data set Frag20-solv-678k, which is developed in this work and contains 678,916 molecular conformations, up to 20 heavy atoms, and their properties calculated at B3LYP/6-31G* level of theory with continuum solvent models, sPhysNet-MT-ens5 predicts density functional theory (DFT)-level electronic energies directly from force field-optimized geometry within chemical accuracy. On the experimental data sets, sPhysNet-MT-ens5 achieves state-of-the-art performances, which predict both experimental hydration free energy with a RMSE of 0.620 kcal/mol on the FreeSolv data set and experimental logP with a RMSE of 0.393 on the PHYSPROP data set. Furthermore, sPhysNet-MT-ens5 also provides a reasonable estimation of model uncertainty which shows correlations with prediction error. Finally, by analyzing the atomic contributions of its predictions, we find that the developed deep learning model is aware of the chemical environment of each atom by assigning reasonable atomic contributions consistent with our chemical knowledge.

Predicting In Vivo Compound Brain Penetration Using Multi-task Graph Neural Networks

Assessing whether compounds penetrate the brain can become critical in drug discovery, either to prevent adverse events or to reach the biological target. Generally, pre-clinical in vivo studies measuring the ratio of brain and blood concentrations (K_p) are required to estimate the brain penetration potential of a new drug entity. In this work, we developed machine learning models to predict in vivo compound brain penetration (as LogK_p) from chemical structure. Our results show the benefit of including in vitro experimental data as auxiliary tasks in multi-task graph neural network (MT-GNN) models. MT-GNNs outperformed single-task (ST) models solely trained on in vivo brain penetration data. The best-performing MT-GNN regression model achieved a coefficient of determination of 0.42 and a mean absolute error of 0.39 (2.5-fold) on a prospective validation set and outperformed all tested ST models. To facilitate decision-making, compounds were classified into brain-penetrant or non-penetrant, achieving a Matthew's correlation coefficient of 0.66. Taken together, our findings indicate that the inclusion of in vitro assay data as MT-GNN auxiliary tasks improves in vivo brain penetration predictions and prospective compound prioritization.

Modeling of SGLT1 in Reconstituted Systems Reveals Apparent Ion-Dependencies of Glucose Uptake and Strengthens the Notion of Water-Permeable Apo States

The reconstitution of secondary active transporters into liposomes shed light on their molecular transport mechanism. The latter are either symporters, antiporters or exchangers, which use the energy contained in the electrochemical gradient of ions to fuel concentrative uptake of their cognate substrate. In liposomal preparations, these gradients can be set by the experimenter. However, due to passive diffusion of the ions and solutes through the membrane, the gradients are not stable and little is known on the time course by which they dissipate and how the presence of a transporter affects this process. Gradient dissipation can also generate a transmembrane potential (VM). Because it is the effective ion gradient, which together with VM fuels concentrative uptake, knowledge on how these parameters change within the time frame of the conducted experiment is key to understanding experimental outcomes. Here, we addressed this problem by resorting to a modelling approach. To this end, we mathematically modeled the liposome in the assumed presence and absence of the sodium glucose transporter 1 (SGLT1). We show that 1) the model can prevent us from reaching erroneous conclusions on the driving forces of substrate uptake and we 2) demonstrate utility of the model in the assignment of the states of SGLT1, which harbor a water channel.

In silico-in vitro estimation of lipophilicity and permeability association for succinimide derivatives using chromatographic anisotropic systems and parallel artificial membrane permeability assay

Passive permeability is one of the key features that determine absorbability and one of the most studied properties in the early phases of drug development. Newly synthesized succinimide derivatives from two different series (1-aryl-3-methylsuccinimides and 1-aryl-3-ethyl-3-methylsuccinimides) with high biological potential have been subjected to estimation of their passive permeability and their association with (a) experimentally obtained anisotropic lipophilicity, (b) in silico-calculated lipophilicity and (c) in silico-predicted permeability and absorbability. Non-cellular-based parallel artificial membrane permeability assay was applied for quantifying their passive permeation, expressed as logP_app . Passive permeation was governed by the lipophilicity of the analysed compounds, and anisotropic lipophilicity was related with statistically significant passive transcellular diffusion (r²  = 0.614, P < 0.001). Moreover, experimentally determined passive permeability, logP_app , was statistically significantly associated with both in silico-predicted absorption constant, k_a (r²  = 0.7886, P < 0.001), and human intestinal absorption (HIA) in percentage (r²  = 0.484, P < 0.001), respectively. However, there was no statistically significant relationship between experimentally obtained permeability on non-cellular-based model and in silico-predicted Caco-2 permeability based on the predictions conducted on two different software. Based on the obtained results, anisotropic systems are promising surrogates for determining lipophilicity, except for compounds with acidic functional groups that are completely ionized under (pH = 7.4).

Artificial intelligence to bring nanomedicine to life

The technology of drug delivery systems (DDSs) has demonstrated an outstanding performance and effectiveness in production of pharmaceuticals, as it is proved by many FDA-approved nanomedicines that have an enhanced selectivity, manageable drug release kinetics and synergistic therapeutic actions. Nonetheless, to date, the rational design and high-throughput development of nanomaterial-based DDSs for specific purposes is far from a routine practice and is still in its infancy, mainly due to the limitations in scientists' capabilities to effectively acquire, analyze, manage, and comprehend complex and ever-growing sets of experimental data, which is vital to develop DDSs with a set of desired functionalities. At the same time, this task is feasible for the data-driven approaches, high throughput experimentation techniques, process automatization, artificial intelligence (AI) technology, and machine learning (ML) approaches, which is referred to as The Fourth Paradigm of scientific research. Therefore, an integration of these approaches with nanomedicine and nanotechnology can potentially accelerate the rational design and high-throughput development of highly efficient nanoformulated drugs and smart materials with pre-defined functionalities. In this Review, we survey the important results and milestones achieved to date in the application of data science, high throughput, as well as automatization approaches, combined with AI and ML to design and optimize DDSs and related nanomaterials. This manuscript mission is not only to reflect the state-of-art in data-driven nanomedicine, but also show how recent findings in the related fields can transform the nanomedicine's image. We discuss how all these results can be used to boost nanomedicine translation to the clinic, as well as highlight the future directions for the development, data-driven, high throughput experimentation-, and AI-assisted design, as well as the production of nanoformulated drugs and smart materials with pre-defined properties and behavior. This Review will be of high interest to the chemists involved in materials science, nanotechnology, and DDSs development for biomedical applications, although the general nature of the presented approaches enables knowledge translation to many other fields of science.

Rotational Variance-Based Data Augmentation in 3D Graph Convolutional Network

This work proposes the data augmentation by molecular rotation, with consideration that the protein-ligand binding events are rotation-variant. As a proof-of-concept, known active (i. e., 1-labeled) ligands to human β-secretase 1 (BACE-1) are rotated for the generation of 0-labeled data, and the rotation-dependent prediction accuracy of 3D graph convolutional network (3DGCN) is investigated after data augmentation. The data augmentation makes the orientation-recognizing ability of 3DGCN improved significantly in the classification task for BACE-1/ligand binding. Furthermore, the data-augmented 3DGCN has a capability for predicting active ligands from a candidate dataset, via improved performance of orientation recognition, which would be applied to virtual drug screening and discovery.