DFT-Machine Learning Approach for Accurate Prediction of pKa

A QSPR Model for Henry's Law Constants of Organic Compounds in Water and Ethanol for Distilled Spirits

Henry's law describes the vapor-liquid equilibrium for dilute gases dissolved in a liquid solvent phase. Descriptions of vapor-liquid equilibrium allow the design of improved separations in the food and beverage industry. The consumer experience of taste and odor are greatly affected by the liquid and vapor phase behavior of organic compounds. This study presents a machine learning (ML) based model that allows quick, accurate predictions of Henry's law constants (kH) for many common organic compounds. Users input only a Simplified Molecular-Input Line-Entry System (SMILES) string or a common English name, and the model returns Henry's law estimates for compounds in water and ethanol. Training was performed on 5,690 compounds. Training data were gathered from an existing database and were supplemented with quantum mechanical (QM) calculations. An extra trees regression model was generated that predicts kH with a mean absolute error of 1.3 in log space and an R2 of 0.98. The model is applied to common flavor and odor compounds in bourbon whiskey as a test case for food and beverage applications.

Leveraging DFT and Molecular Fragmentation for Chemically Accurate pKa Prediction Using Machine Learning

We present a quantum mechanical/machine learning (ML) framework based on random forest to accurately predict the pKas of complex organic molecules using inexpensive density functional theory (DFT) calculations. By including physics-based features from low-level DFT calculations and structural features from our connectivity-based hierarchy (CBH) fragmentation protocol, we can correct the systematic error associated with DFT. The generalizability and performance of our model are evaluated on two benchmark sets (SAMPL6 and Novartis). We believe the carefully curated input of physics-based features lessens the model's data dependence and need for complex deep learning architectures, without compromising the accuracy of the test sets. As a point of novelty, our work extends the applicability of CBH, employing it for the generation of viable molecular descriptors for ML.

Building Chemical Intuition about Physicochemical Properties of C8-Per-/Polyfluoroalkyl Carboxylic Acids through Computational Means

We have predicted acid dissociation constants (pK a), octanol-water partition coefficients (K OW), and DMPC lipid membrane-water partition coefficients (K lipid-w) of 150 different eight-carbon-containing poly-/perfluoroalkyl carboxylic acids (C8-PFCAs) utilizing the COnductor-like Screening MOdel for Realistic Solvents (COSMO-RS) theory. Different trends associated with functionalization, degree of fluorination, degree of saturation, degree of chlorination, and branching are discussed on the basis of the predicted values for the partition coefficients. In general, functionalization closest to the carboxylic headgroup had the greatest impact on the value of the predicted physicochemical properties.

Accurate Computation of Aqueous pKas of Biologically Relevant Organic Acids: Overcoming the Challenges Posed by Multiple Conformers, Tautomeric Equilibria, and Disparate Functional Groups with the Fully Black-Box pK-Yay Method

The use of static electronic structure calculations to compute solution-phase pKas offers a great advantage in that a macroscopic bulk property could be computed via microscopic computations involving very few molecules. There are various sources of errors in the quantum chemical calculations though. Overcoming these errors to accurately compute pKas of a plethora of acids is an active area of research in physical chemistry pursued by both computational as well as experimental chemists. We recently developed the pK-Yay method in our attempt to accurately compute aqueous pKas of strong and weak acids. The method is fully black-box, computationally inexpensive, and is very easy for even a nonexpert to use. However, the method was thus far tested on very few molecules (only 16 in all). Herein, in order to assess the future applicability of pK-Yay, we study the effect of multiple conformers, the presence of tautomers under equilibrium, and the impact of a wide variety of functional groups (derivatives of acetic acid with substituents at various positions, dicarboxylic acids, aromatic carboxylic acids, amines and amides, phenols and thiols, and fluorine bearing organic acids). Starting with more than 1000 conformers and tautomers, this study establishes that overall errors of ∼ 1.0 pKa units are routinely obtained for a majority of the molecules. Larger errors are noted in cases where multiple charges, intramolecular hydrogen bonding, and several ionizable functional groups are simultaneously present. An important conclusion to emerge from this work is that, the computed pKas are insensitive (difference <0.5) to whether we consider multiple conformers/tautomers or only choose the most stable conformer/tautomer. Further, pK-Yay captures the stereoelectronic effects arising due to differing axial vs equatorial pattern, and is useful to predict the dominant acid-base equilibrium in a system featuring several equilibria. Overall, pK-Yay may be employed in several chemical applications featuring organic molecules and biomonomers.

Spectroscopic characterization, molecular docking and machine learning studies of sulphur containing hydrazide derivatives

Machine learning applied in chemistry is a growing field of research. For assessing structure-property variations, this paper describes in silico studies of the hydrazide derivatives of thiosemicarbazide (TSCZ) and thiocarbohydrazide (TCHZ). The structures of TSCZ and TCHZ have been elucidated using modern spectroscopic techniques. The UV-vis spectra showed strong charge transfer transitions (π-π*) for TSCZ and TCHZ with high extinction coefficients. The NBO analysis showed orbital overlap between lp1 (N2) and σ* (C3-S4) in TSCZ and TCHZ due to intramolecular charge transfer. The first hyperpolarizabilities (β0) for TSCZ and TCHZ were found to be 0.7155 and 2.1615 × 10-30 esu, respectively, indicating their greater suitability for NLO applications as compared to standard reference urea. The strong electrophilic behaviour of TSCZ and TCHZ has been indicated by their global elecrophilicity index. The electrophilic reactivity descriptor analysis indicated that the investigated molecules could serve as precursors for the targeted synthesis of new heterocyclic derivatives. The docking studies showed appreciable binding energies with target proteins having PDB IDs 2WJE and 6CLU of Gram-positive bacteria, namely, Streptococcus pneumoniae phosphatase (PTP-CPS4B) and Staphylococcus aureus dihydropteroate synthase (saDHPS), respectively, for TSCZ and TCHZ, predicting good antimicrobial activity.

Calculating the Aqueous pKa of Phenols: Predictions for Antioxidants and Cannabinoids

We aim to develop a theoretical methodology for the accurate aqueous pK_a prediction of structurally complex phenolic antioxidants and cannabinoids. In this study, five functionals (M06-2X, B3LYP, BHandHLYP, PBE0, and TPSS) and two solvent models (SMD and PCM) were combined with the 6-311++G(d,p) basis set to predict pK_a values for twenty structurally simple phenols. None of the direct calculations produced good results. However, the correlations between the calculated Gibbs energy difference of each acid and its conjugate base, ΔGaq(BA)°=ΔGaqA-°-ΔGaq(HA)°, and the experimental aqueous pK_a values had superior predictive accuracy, which was also tested relative to an independent set of ten molecules of which six were structurally complex phenols. New correlations were built with twenty-seven phenols (including the phenols with experimental pK_a values from the test set), which were used to make predictions. The best correlation equations used the PCM method and produced mean absolute errors of 0.26-0.27 pK_a units and R² values of 0.957-0.960. The average range of predictions for the potential antioxidants (cannabinoids) was 0.15 (0.25) pK_a units, which indicates good agreement between our methodologies. The new correlation equations could be used to make pK_a predictions for other phenols in water and potentially in other solvents where they might be more soluble.

A Guide to In Silico Drug Design

The drug discovery process is a rocky path that is full of challenges, with the result that very few candidates progress from hit compound to a commercially available product, often due to factors, such as poor binding affinity, off-target effects, or physicochemical properties, such as solubility or stability. This process is further complicated by high research and development costs and time requirements. It is thus important to optimise every step of the process in order to maximise the chances of success. As a result of the recent advancements in computer power and technology, computer-aided drug design (CADD) has become an integral part of modern drug discovery to guide and accelerate the process. In this review, we present an overview of the important CADD methods and applications, such as in silico structure prediction, refinement, modelling and target validation, that are commonly used in this area.

Machine learning methods for pKa prediction of small molecules: Advances and challenges

The acid-base dissociation constant (pK_a) is a fundamental property influencing many ADMET properties of small molecules. However, rapid and accurate pK_a prediction remains a great challenge. In this review, we outline the current advances in machine-learning-based QSAR models for pK_a prediction, including descriptor-based and graph-based approaches, and summarize their pros and cons. Moreover, we highlight the current challenges and future directions regarding experimental data, crucial factors influencing pK_a and in silico prediction tools. We hope that this review can provide a practical guidance for the follow-up studies.

Accelerating Solvent Selection for Type II Porous Liquids

Type II porous liquids, comprising intrinsically porous molecules dissolved in a liquid solvent, potentially combine the adsorption properties of porous adsorbents with the handling advantages of liquids. Previously, discovery of appropriate solvents to make porous liquids had been limited to direct experimental tests. We demonstrate an efficient screening approach for this task that uses COSMO-RS calculations, predictions of solvent pK_a values from a machine-learning model, and several other features and apply this approach to select solvents from a library of more than 11,000 compounds. This method is shown to give qualitative agreement with experimental observations for two molecular cages, CC13 and TG-TFB-CHEDA, identifying solvents with higher solubility for these molecules than had previously been known. Ultimately, the algorithm streamlines the downselection of suitable solvents for porous organic cages to enable more rapid discovery of Type II porous liquids.