Machine learning in chemical reaction space

Predicting Emission Spectra of Heteroleptic Iridium Complexes Using Artificial Chemical Intelligence

We report a deep learning-based approach to accurately predict the emission spectra of phosphorescent heteroleptic [Ir(C ∧ N ${{\rm{C}}^\wedge {\rm{N}}}$ )2(N ∧ N ${{\rm{N}}^\wedge {\rm{N}}}$ )]+ complexes, enabling the rapid discovery of novel Ir(III) chromophores for diverse applications including organic light-emitting diodes and solar fuel cells. The deep learning models utilize graph neural networks and other chemical features in architectures that reflect the inherent structure of the heteroleptic complexes, composed ofC ∧ N ${{\rm{C}}^\wedge {\rm{N}}}$ andN ∧ N ${{\rm{N}}^\wedge {\rm{N}}}$ ligands, and are thus geared towards efficient training over the dataset. By leveraging experimental emission data, our models reliably predict the full emission spectra of these complexes across various emission profiles, surpassing the accuracy of conventional DFT and correlated wavefunction methods, while simultaneously achieving robustness to the presence of imperfect (noisy, low-quality) training spectra. We showcase the potential applications for these and related models for in silico prediction of complexes with tailored emission properties, as well as in "design of experiment" contexts to reduce the synthetic burden of high-throughput screening. In the latter case, we demonstrate that the models allow us to exploit a limited amount of experimental data to explore a wide range of chemical space, thus leveraging a modest synthetic effort.

Superior electroadhesion force with permittivity-engineered bilayer films using electrostatic simulation and machine learning approaches

Electroadhesive forces are crucial in various applications, including grasping devices, electro-sticky boards, electrostatic levitation, and climbing robots. However, the design of electroadhesive devices relies on speculative or empirical error approaches. Therefore, we present a theoretical model comprising predictive coplanar electrodes and protective layers for analyzing the electrostatic fields between an object and electroadhesive device. The model considers the role of protective layer and the air gap between the electrode surface and the object. To exert a higher electroadhesive force, the higher permeability of the protective layer is required. However, a high permeability of the protective layer is hard to withstand high applied voltage. To overcome this, two materials with different permeabilities were employed as protective layers-a low-permeability inner layer and a high-permeability outer layer-to maintain a high voltage and generate a large electroadhesive force. Because a low-permeability inner layer material was selected, a more permeable outer layer material was considered. A theoretical analysis revealed complex relationships between various design parameters. The impact of key design parameters and working environments on the electroadhesion behavior was further investigated. This study reveals the fundamental principles of electroadhesion and proposes prospective methods to enhance the design of electroadhesive devices for various engineering applications.

Self-assembly prediction of architecture-controlled bottlebrush copolymers in solution using graph convolutional networks

The investigation of bottlebrush copolymer self-assembly in solution involves a comprehensive approach integrating simulation and experimental research, due to their unique physical characteristics. However, the intricate architecture of bottlebrush copolymers and the diverse solvent conditions introduce a wide range of parameter spaces. In this study, we investigated the solution self-assembly behavior of bottlebrush copolymers by combining dissipative particle dynamics (DPD) simulation results and machine learning (ML) including graph convolutional networks (GCNs). The architecture of bottlebrush copolymers is encoded by graphs including connectivity, side chain length, bead types, and interaction parameters of DPD simulation. Using GCN, we accurately predicted the single chain properties of bottlebrush copolymers with over 95% accuracy. Furthermore, phase behavior was precisely predicted using these single chain properties. Shapley additive explanations (SHAP) values of single chain properties to the various self-assembly morphologies were calculated to investigate the correlation between single chain properties and morphologies. In addition, we analyzed single chain properties and phase behavior as a function of DPD interaction parameters, extracting relevant physical properties for vesicle morphology formation. This work paves the way for tailored design in solution of self-assembled nanostructures of bottlebrush copolymers, offering a GCN framework for precise prediction of self-assembly morphologies under various chain architectures and solvent conditions.

Identification of Reaction Network Hypotheses for Complex Feedstocks from Spectroscopic Measurements with Minimal Human Intervention

In this work, we detail an automated reaction network hypothesis generation protocol for processes involving complex feedstocks where information about the species and reactions involved is unknown. Our methodology is process agnostic and can be utilized in any reactive process with spectroscopic measurements that provide information on the evolution of the components in the mixture. We decompose the mixture spectra to obtain spectroscopic signatures of the individual components and use a 1-D convolutional neural network to automatically identify functional groups indicated by them. We employ atom-atom mapping to automatically recover reaction rules that are applied on candidate molecules identified from chemistry databases through fingerprint similarity. The method is tested on synthetic data and on spectroscopic measurements of lab-scale batch hydrothermal liquefaction (HTL) of biomass to determine the accuracy of prediction across datasets of varying complexities. Our methodology is able to identify reaction network hypotheses containing reaction networks close to the ground truth in the case of synthetic data, and we are also able to recover candidate molecules and reaction networks close to the ones reported in the previous literature studies for biomass pyrolysis.

SchNetPack Hyperparameter Optimization for a More Reliable Top Docking Scores Prediction

Options to improve the extrapolation power of the neural network designed using the SchNetPack package with respect to top docking scores prediction are presented. It is shown that hyperparameter tuning of the atomistic model representation (in the schnetpack.representation) improves the prediction of the top scoring compounds, which have characteristically a low incidence in randomized data sets for training of machine learning models. The prediction robustness is evaluated according to the mean square error (MSE) and the entropy of the average loss landscape decrease. Admittedly, the improvement of the top scoring compounds' prediction accuracy comes with the penalty of worsening the overall prediction power. It is revealed that the most impactful hyperparameter is the cutoff (5 Å is reported as the optimal choice). Other parameters (e.g., number of radial basis functions, number of interaction layers of the neural network, feature vector size or its batch size) are found to not affect the prediction robustness of the top scoring compounds in any comparable way relative to the cutoff. The MSE of the best docking score prediction (below -13 kcal/mol) improves from ca. 3.5 to 0.9 kcal/mol, while the prediction of less potent compounds (-13 to -11 kcal/mol) shows a lesser improvement, i.e., a decrease of MSE from 1.6 to 1.3 kcal/mol. Additionally, oversampling and undersampling of the training set with respect to the top scoring compounds' abundance is presented. The results indicate that the cutoff choice performs better than over- or undersampling of the training set, with undersampling performing better than oversampling.

Comparing ANI-2x, ANI-1ccx neural networks, force field, and DFT methods for predicting conformational potential energy of organic molecules

In this study, the conformational potential energy surfaces of Amylmetacresol, Benzocaine, Dopamine, Betazole, and Betahistine molecules were scanned and analyzed using the neural network architecture ANI-2 × and ANI-1ccx, the force field method OPLS, and density functional theory with the exchange-correlation functional B3LYP and the basis set 6-31G(d). The ANI-1ccx and ANI-2 × methods demonstrated the highest accuracy in predicting torsional energy profiles, effectively capturing the minimum and maximum values of these profiles. Conformational potential energy values calculated by B3LYP and the OPLS force field method differ from those calculated by ANI-1ccx and ANI-2x, which account for non-bonded intramolecular interactions, since the B3LYP functional and OPLS force field weakly consider van der Waals and other intramolecular forces in torsional energy profiles. For a more comprehensive analysis, electronic parameters such as dipole moment, HOMO, and LUMO energies for different torsional angles were calculated at two levels of theory, B3LYP/6-31G(d) and ωB97X/6-31G(d). These calculations confirmed that ANI predictions are more accurate than density functional theory calculations with B3LYP functional and OPLS force field for determining potential energy surfaces. This research successfully addressed the challenges in determining conformational potential energy levels and shows how machine learning and deep neural networks offer a more accurate, cost-effective, and rapid alternative for predicting torsional energy profiles.

Benchmarking machine-readable vectors of chemical reactions on computed activation barriers

In recent years, there has been a surge of interest in predicting computed activation barriers, to enable the acceleration of the automated exploration of reaction networks. Consequently, various predictive approaches have emerged, ranging from graph-based models to methods based on the three-dimensional structure of reactants and products. In tandem, many representations have been developed to predict experimental targets, which may hold promise for barrier prediction as well. Here, we bring together all of these efforts and benchmark various methods (Morgan fingerprints, the DRFP, the CGR representation-based Chemprop, SLATMd, B2Rl2, EquiReact and language model BERT + RXNFP) for the prediction of computed activation barriers on three diverse datasets.

Machine-Learned Kohn-Sham Hamiltonian Mapping for Nonadiabatic Molecular Dynamics

In this work, we report a simple, efficient, and scalable machine-learning (ML) approach for mapping non-self-consistent Kohn-Sham Hamiltonians constructed with one kind of density functional to the nearly self-consistent Hamiltonians constructed with another kind of density functional. This approach is designed as a fast surrogate Hamiltonian calculator for use in long nonadiabatic dynamics simulations of large atomistic systems. In this approach, the input and output features are Hamiltonian matrices computed from different levels of theory. We demonstrate that the developed ML-based Hamiltonian mapping method (1) speeds up the calculations by several orders of magnitude, (2) is conceptually simpler than alternative ML approaches, (3) is applicable to different systems and sizes and can be used for mapping Hamiltonians constructed with arbitrary density functionals, (4) requires a modest training data, learns fast, and generates molecular orbitals and their energies with the accuracy nearly matching that of conventional calculations, and (5) when applied to nonadiabatic dynamics simulation of excitation energy relaxation in large systems yields the corresponding time scales within the margin of error of the conventional calculations. Using this approach, we explore the excitation energy relaxation in C60 fullerene and Si75H64 quantum dot structures and derive qualitative and quantitative insights into dynamics in these systems.

Exploring Chemical Space Using Ab Initio Hyperreactor Dynamics

In recent years, first-principles exploration of chemical reaction space has provided valuable insights into intricate reaction networks. Here, we introduce ab initio hyperreactor dynamics, which enables rapid screening of the accessible chemical space from a given set of initial molecular species, predicting new synthetic routes that can potentially guide subsequent experimental studies. For this purpose, different hyperdynamics derived bias potentials are applied along with pressure-inducing spherical confinement of the molecular system in ab initio molecular dynamics simulations to efficiently enhance reactivity under mild conditions. To showcase the advantages and flexibility of the hyperreactor approach, we present a systematic study of the method's parameters on a HCN toy model and apply it to a recently introduced experimental model for the prebiotic formation of glycinal and acetamide in interstellar ices, which yields results in line with experimental findings. In addition, we show how the developed framework enables the study of complicated transitions like the first step of a nonenzymatic DNA nucleoside synthesis in an aqueous environment, where the molecular fragmentation problem of earlier nanoreactor approaches is avoided.

Prediction of chemical reaction yields with large-scale multi-view pre-training

Developing machine learning models with high generalization capability for predicting chemical reaction yields is of significant interest and importance. The efficacy of such models depends heavily on the representation of chemical reactions, which has commonly been learned from SMILES or graphs of molecules using deep neural networks. However, the progression of chemical reactions is inherently determined by the molecular 3D geometric properties, which have been recently highlighted as crucial features in accurately predicting molecular properties and chemical reactions. Additionally, large-scale pre-training has been shown to be essential in enhancing the generalization capability of complex deep learning models. Based on these considerations, we propose the Reaction Multi-View Pre-training (ReaMVP) framework, which leverages self-supervised learning techniques and a two-stage pre-training strategy to predict chemical reaction yields. By incorporating multi-view learning with 3D geometric information, ReaMVP achieves state-of-the-art performance on two benchmark datasets. Notably, the experimental results indicate that ReaMVP has a significant advantage in predicting out-of-sample data, suggesting an enhanced generalization ability to predict new reactions. Scientific Contribution: This study presents the ReaMVP framework, which improves the generalization capability of machine learning models for predicting chemical reaction yields. By integrating sequential and geometric views and leveraging self-supervised learning techniques with a two-stage pre-training strategy, ReaMVP achieves state-of-the-art performance on benchmark datasets. The framework demonstrates superior predictive ability for out-of-sample data and enhances the prediction of new reactions.

Generating a skeleton reaction network for reactions of large-scale ReaxFF MD pyrolysis simulations based on a machine learning predicted reaction class

The reactive molecular dynamics using ReaxFF provides an effective means to generate global reactions for pyrolysis of realistic fuel mixtures. The reactions from large-scale pyrolysis simulations of a fuel mixture may be characterized by multiple reaction sites, explosion of intermediate species structures, and scattered contribution of diversified pathways to product species. This work proposes an approach of SRG-Reax aiming at generating skeleton reaction networks based on reaction patterns or classes of reaction centers from huge reactions obtained from ReaxFF MD simulations of realistic fuel pyrolysis. SRG-Reax (Skeleton Reaction network Generation for ReaxFF MD) is implemented through building a semi-supervised machine learning model of tri-training for predicting the reaction classes of pyrolysis reactions based on an extended reaction center. Three different reaction center descriptions of reaction features and reaction transformation fingerprints are employed as inputs for developing the tri-training classifier. Major reaction pathways can be identified based on reaction class ratios and product species ratios calculated by merging reaction pathways of the same reaction class. The SRG-Reax approach was applied in skeleton reaction network generation for RP-3 pyrolysis based on the ReaxFF MD simulations of a high-fidelity 45-component RP-3 fuel model. The skeleton reaction networks for n-paraffins, iso-paraffins, cycloparaffins, olefins, and aromatics in RP-3 pyrolysis were obtained. The reaction class ratios and product species ratios in the obtained skeleton reaction network provide comprehensive intuitive insight into global pyrolysis chemistry. SRG-Reax has the potential to obtain relatively complete skeleton reaction networks for the pyrolysis of hydrocarbon fuel, polymers, biomass, coal, and more.

Transferable Machine Learning Interatomic Potential for Bond Dissociation Energy Prediction of Drug-like Molecules

We present a transferable MACE interatomic potential that is applicable to open- and closed-shell drug-like molecules containing hydrogen, carbon, and oxygen atoms. Including an accurate description of radical species extends the scope of possible applications to bond dissociation energy (BDE) prediction, for example, in the context of cytochrome P450 (CYP) metabolism. The transferability of the MACE potential was validated on the COMP6 data set, containing only closed-shell molecules, where it reaches better accuracy than the readily available general ANI-2x potential. MACE achieves similar accuracy on two CYP metabolism-specific data sets, which include open- and closed-shell structures. This model enables us to calculate the aliphatic C-H BDE, which allows us to compare reaction energies of hydrogen abstraction, which is the rate-limiting step of the aliphatic hydroxylation reaction catalyzed by CYPs. On the "CYP 3A4" data set, MACE achieves a BDE RMSE of 1.37 kcal/mol and better prediction of BDE ranks than alternatives: the semiempirical AM1 and GFN2-xTB methods and the ALFABET model that directly predicts bond dissociation enthalpies. Finally, we highlight the smoothness of the MACE potential over paths of sp3C-H bond elongation and show that a minimal extension is enough for the MACE model to start finding reasonable minimum energy paths of methoxy radical-mediated hydrogen abstraction. Altogether, this work lays the ground for further extensions of scope in terms of chemical elements, (CYP-mediated) reaction classes and modeling the full reaction paths, not only BDEs.

Multi-level advances in databases related to systems pharmacology in traditional Chinese medicine: a 60-year review

The therapeutic effects of traditional Chinese medicine (TCM) involve intricate interactions among multiple components and targets. Currently, computational approaches play a pivotal role in simulating various pharmacological processes of TCM. The application of network analysis in TCM research has provided an effective means to explain the pharmacological mechanisms underlying the actions of herbs or formulas through the lens of biological network analysis. Along with the advances of network analysis, computational science has coalesced around the core chain of TCM research: formula-herb-component-target-phenotype-ZHENG, facilitating the accumulation and organization of the extensive TCM-related data and the establishment of relevant databases. Nonetheless, recent years have witnessed a tendency toward homogeneity in the development and application of these databases. Advancements in computational technologies, including deep learning and foundation model, have propelled the exploration and modeling of intricate systems into a new phase, potentially heralding a new era. This review aims to delves into the progress made in databases related to six key entities: formula, herb, component, target, phenotype, and ZHENG. Systematically discussions on the commonalities and disparities among various database types were presented. In addition, the review raised the issue of research bottleneck in TCM computational pharmacology and envisions the forthcoming directions of computational research within the realm of TCM.

Machine Learning for Bridging the Gap between Density Functional Theory and Coupled Cluster Energies

Accurate electronic energies and properties are crucial for successful reaction design and mechanistic investigations. Computing energies and properties of molecular structures has proven extremely useful, and, with increasing computational power, the limits of high-level approaches (such as coupled cluster theory) are expanding to ever larger systems. However, because scaling is highly unfavorable, these methods are still not universally applicable to larger systems. To address the need for fast and accurate electronic energies of larger systems, we created a database of around 8000 small organic monomers (2000 dimers) optimized at the B3LYP-D3(BJ)/cc-pVTZ level of theory. This database also includes single-point energies computed at various levels of theory, including PBE1PBE, ωΒ97Χ, M06-2X, revTPSS, B3LYP, and BP86, for density functional theory as well as DLPNO-CCSD(T) and CCSD(T) for coupled cluster theory, all in conjunction with a cc-pVTZ basis. We used this database to train machine learning models based on graph neural networks using two different graph representations. Our models are able to make energy predictions from B3LYP-D3(BJ)/cc-pVTZ inputs to CCSD(T)/cc-pVTZ outputs with a mean absolute error of 0.78 and to DLPNO-CCSD(T)/cc-pVTZ with an mean absolute error of 0.50 and 0.18 kcal mol^-1 for monomers and dimers, respectively. The model for dimers was further validated on the S22 database, and the monomer model was tested on challenging systems, including those with highly conjugated or functionally complex molecules.

Designing main-group catalysts for low-temperature methane combustion by ozone

The catalytic combustion of methane at a low temperature is becoming increasingly key to controlling unburned CH₄ emissions from natural gas vehicles and power plants, although the low activity of benchmark platinum-group-metal catalysts hinders its broad application. Based on automated reaction route mapping, we explore main-group elements catalysts containing Si and Al for low-temperature CH₄ combustion with ozone. Computational screening of the active site predicts that strong Brønsted acid sites are promising for methane combustion. We experimentally demonstrate that catalysts containing strong Bronsted acid sites exhibit improved CH₄ conversion at 250 °C, correlating with the theoretical predictions. The main-group catalyst (proton-type beta zeolite) delivered a reaction rate that is 442 times higher than that of a benchmark catalyst (5 wt% Pd-loaded Al₂O₃) at 190 °C and exhibits higher tolerance to steam and SO₂. Our strategy demonstrates the rational design of earth-abundant catalysts based on automated reaction route mapping.

The application of machine learning to air pollution research: A bibliometric analysis

Machine learning (ML) is an advanced computer algorithm that simulates the human learning process to solve problems. With an explosion of monitoring data and the increasing demand for fast and accurate prediction, ML models have been rapidly developed and applied in air pollution research. In order to explore the status of ML applications in air pollution research, a bibliometric analysis was made based on 2962 articles published from 1990 to 2021. The number of publications increased sharply after 2017, comprising approximately 75% of the total. Institutions in China and United States contributed half of all publications with most research being conducted by individual groups rather than global collaborations. Cluster analysis revealed four main research topics for the application of ML: chemical characterization of pollutants, short-term forecasting, detection improvement and optimizing emission control. The rapid development of ML algorithms has increased the capability to explore the chemical characteristics of multiple pollutants, analyze chemical reactions and their driving factors, and simulate scenarios. Combined with multi-field data, ML models are a powerful tool for analyzing atmospheric chemical processes and evaluating the management of air quality and deserve greater attention in future.

Physics-inspired machine learning of localized intensive properties

Machine learning (ML) has been widely applied to chemical property prediction, most prominently for the energies and forces in molecules and materials. The strong interest in predicting energies in particular has led to a 'local energy'-based paradigm for modern atomistic ML models, which ensures size-extensivity and a linear scaling of computational cost with system size. However, many electronic properties (such as excitation energies or ionization energies) do not necessarily scale linearly with system size and may even be spatially localized. Using size-extensive models in these cases can lead to large errors. In this work, we explore different strategies for learning intensive and localized properties, using HOMO energies in organic molecules as a representative test case. In particular, we analyze the pooling functions that atomistic neural networks use to predict molecular properties, and suggest an orbital weighted average (OWA) approach that enables the accurate prediction of orbital energies and locations.

Combustion, Chemistry, and Carbon Neutrality

Combustion is a reactive oxidation process that releases energy bound in chemical compounds used as fuels─energy that is needed for power generation, transportation, heating, and industrial purposes. Because of greenhouse gas and local pollutant emissions associated with fossil fuels, combustion science and applications are challenged to abandon conventional pathways and to adapt toward the demand of future carbon neutrality. For the design of efficient, low-emission processes, understanding the details of the relevant chemical transformations is essential. Comprehensive knowledge gained from decades of fossil-fuel combustion research includes general principles for establishing and validating reaction mechanisms and process models, relying on both theory and experiments with a suite of analytic monitoring and sensing techniques. Such knowledge can be advantageously applied and extended to configure, analyze, and control new systems using different, nonfossil, potentially zero-carbon fuels. Understanding the impact of combustion and its links with chemistry needs some background. The introduction therefore combines information on exemplary cultural and technological achievements using combustion and on nature and effects of combustion emissions. Subsequently, the methodology of combustion chemistry research is described. A major part is devoted to fuels, followed by a discussion of selected combustion applications, illustrating the chemical information needed for the future.

Neural network potentials for accelerated metadynamics of oxygen reduction kinetics at Au-water interfaces

The application of ab initio molecular dynamics (AIMD) for the explicit modeling of reactions at solid-liquid interfaces in electrochemical energy conversion systems like batteries and fuel cells can provide new understandings towards reaction mechanisms. However, its prohibitive computational cost severely restricts the time- and length-scales of AIMD. Equivariant graph neural network (GNN) based accurate surrogate potentials can accelerate the speed of performing molecular dynamics after learning on representative structures in a data efficient manner. In this study, we combined uncertainty-aware GNN potentials and enhanced sampling to investigate the reactive process of the oxygen reduction reaction (ORR) at an Au(100)-water interface. By using a well-established active learning framework based on CUR matrix decomposition, we can evenly sample equilibrium structures from MD simulations and non-equilibrium reaction intermediates that are rarely visited during the reaction. The trained GNNs have shown exceptional performance in terms of force prediction accuracy, the ability to reproduce structural properties, and low uncertainties when performing MD and metadynamics simulations. Furthermore, the collective variables employed in this work enabled the automatic search of reaction pathways and provide a detailed understanding towards the ORR reaction mechanism on Au(100). Our simulations identified the associative reaction mechanism without the presence of *O and a low reaction barrier of 0.3 eV, which is in agreement with experimental findings. The methodology employed in this study can pave the way for modeling complex chemical reactions at electrochemical interfaces with an explicit solvent under ambient conditions.

An automated reaction route mapping for the reaction of NO and active species on Ag4 clusters in zeolites

A computational investigation of the catalytic reaction on multinuclear sites is very challenging. Here, using an automated reaction route mapping method, the single-component artificial force induced reaction (SC-AFIR) algorithm, the catalytic reaction of NO and OH/OOH species over the Ag₄²⁺ cluster in a zeolite is investigated. The results of the reaction route mapping for H₂ + O₂ reveal that OH and OOH species are formed over the Ag₄²⁺ cluster via an activation barrier lower than that of OH formation from H₂O dissociation. Then, reaction route mapping is performed to examine the reactivity of the OH and OOH species with NO molecules over the Ag₄²⁺ cluster, resulting in the facile reaction path of HONO formation. With the aid of the automated reaction route mapping, the promotion effect of H₂ addition on the SCR reaction was computationally proposed (boosting the formation of OH and OOH species). In addition, the present study emphasizes that automated reaction route mapping is a powerful tool to elucidate the complicated reaction pathway on multi-nuclear clusters.

Benchmark of general-purpose machine learning-based quantum mechanical method AIQM1 on reaction barrier heights

Artificial intelligence-enhanced quantum mechanical method 1 (AIQM1) is a general-purpose method that was shown to achieve high accuracy for many applications with a speed close to its baseline semiempirical quantum mechanical (SQM) method ODM2*. Here, we evaluate the hitherto unknown performance of out-of-the-box AIQM1 without any refitting for reaction barrier heights on eight datasets, including a total of ∼24 thousand reactions. This evaluation shows that AIQM1's accuracy strongly depends on the type of transition state and ranges from excellent for rotation barriers to poor for, e.g., pericyclic reactions. AIQM1 clearly outperforms its baseline ODM2* method and, even more so, a popular universal potential, ANI-1ccx. Overall, however, AIQM1 accuracy largely remains similar to SQM methods (and B3LYP/6-31G* for most reaction types) suggesting that it is desirable to focus on improving AIQM1 performance for barrier heights in the future. We also show that the built-in uncertainty quantification helps in identifying confident predictions. The accuracy of confident AIQM1 predictions is approaching the level of popular density functional theory methods for most reaction types. Encouragingly, AIQM1 is rather robust for transition state optimizations, even for the type of reactions it struggles with the most. Single-point calculations with high-level methods on AIQM1-optimized geometries can be used to significantly improve barrier heights, which cannot be said for its baseline ODM2* method.

Benzoquinoline Chemical Space: A Helpful Approach in Antibacterial and Anticancer Drug Design

Benzoquinolines are used in many drug design projects as starting molecules subject to derivatization. This computational study aims to characterize e benzoquinone drug space to ease future drug design processes based on these molecules. The drug space is composed of all benzoquinones, which are active on topoisomerase II and ATP synthase. Topological, chemical, and bioactivity spaces are explored using computational methodologies based on virtual screening and scaffold hopping and molecular docking, respectively. Topological space is a geometrical space in which the elements composing it can be defined as a set of neighbors (which satisfy a particular axiom). In such space, a chemical space can be defined as the property space spanned by all possible molecules and chemical compounds adhering to a given set of construction principles and boundary conditions. In this chemical space, the potentially pharmacologically active molecules form the bioactivity space. Results show a poly-morphological chemical space that suggests distinct characteristics. The chemical space is correlated with properties such as steric energy, the number of hydrogen bonds, the presence of halogen atoms, and membrane permeability-related properties. Lastly, novel chemical compounds (such as oxadiazole methybenzamide and floro methylcyclohexane diene) with drug-like potential, active on TOPO II and ATP synthase have been identified.

Chemical reaction networks and opportunities for machine learning

Chemical reaction networks (CRNs), defined by sets of species and possible reactions between them, are widely used to interrogate chemical systems. To capture increasingly complex phenomena, CRNs can be leveraged alongside data-driven methods and machine learning (ML). In this Perspective, we assess the diverse strategies available for CRN construction and analysis in pursuit of a wide range of scientific goals, discuss ML techniques currently being applied to CRNs and outline future CRN-ML approaches, presenting scientific and technical challenges to overcome.

An automated method for graph-based chemical space exploration and transition state finding

Algorithms that automatically explore the chemical space have been limited to chemical systems with a low number of atoms due to expensive involved quantum calculations and the large amount of possible reaction pathways. The method described here presents a novel solution to the problem of chemical exploration by generating reaction networks with heuristics based on chemical theory. First, a second version of the reaction network is determined through molecular graph transformations acting upon functional groups of the reacting. Only transformations that break two chemical bonds and form two new ones are considered, leading to a significant performance enhancement compared to previously presented algorithm. Second, energy barriers for this reaction network are estimated through quantum chemical calculations by a growing string method, which can also identify non-octet species missed during the previous step and further define the reaction network. The proposed algorithm has been successfully applied to five different chemical reactions, in all cases identifying the most important reaction pathways.

Graph-Driven Reaction Discovery: Progress, Challenges, and Future Opportunities

Graph-based descriptors, such as bond-order matrices and adjacency matrices, offer a simple and compact way of categorizing molecular structures; furthermore, such descriptors can be readily used to catalog chemical reactions (i.e., bond-making and -breaking). As such, a number of graph-based methodologies have been developed with the goal of automating the process of generating chemical reaction network models describing the possible mechanistic chemistry in a given set of reactant species. Here, we outline the evolution of these graph-based reaction discovery schemes, with particular emphasis on more recent methods incorporating graph-based methods with semiempirical and ab initio electronic structure calculations, minimum-energy path refinements, and transition state searches. Using representative examples from homogeneous catalysis and interstellar chemistry, we highlight how these schemes increasingly act as "virtual reaction vessels" for interrogating mechanistic questions. Finally, we highlight where challenges remain, including issues of chemical accuracy and calculation speeds, as well as the inherent challenge of dealing with the vast size of accessible chemical reaction space.

Machine Learning for Chemical Reactivity: The Importance of Failed Experiments

Assessing the outcomes of chemical reactions in a quantitative fashion has been a cornerstone across all synthetic disciplines. Classically approached through empirical optimization, data-driven modelling bears an enormous potential to streamline this process. However, such predictive models require significant quantities of high-quality data, the availability of which is limited: Main reasons for this include experimental errors and, importantly, human biases regarding experiment selection and result reporting. In a series of case studies, we investigate the impact of these biases for drawing general conclusions from chemical reaction data, revealing the utmost importance of "negative" examples. Eventually, case studies into data expansion approaches showcase directions to circumvent these limitations-and demonstrate perspectives towards a long-term data quality enhancement in chemistry.

Low-cost prediction of molecular and transition state partition functions via machine learning

We have generated an open-source dataset of over 30 000 organic chemistry gas phase partition functions. With this data, a machine learning deep neural network estimator was trained to predict partition functions of unknown organic chemistry gas phase transition states. This estimator only relies on reactant and product geometries and partition functions. A second machine learning deep neural network was trained to predict partition functions of chemical species from their geometry. Our models accurately predict the logarithm of test set partition functions with a maximum mean absolute error of 2.7%. Thus, this approach provides a means to reduce the cost of computing reaction rate constants ab initio. The models were also used to compute transition state theory reaction rate constant prefactors and the results were in quantitative agreement with the corresponding ab initio calculations with an accuracy of 98.3% on the log scale.

Successes and challenges in using machine-learned activation energies in kinetic simulations

The prediction of the thermodynamic and kinetic properties of chemical reactions is increasingly being addressed by machine-learning (ML) methods such as artificial neural networks (ANNs). While a number of recent studies have reported success in predicting chemical reaction activation energies, less attention has focused on how the accuracy of ML predictions filter through to predictions of macroscopic observables. Here, we consider the impact of the uncertainty associated with ML prediction of activation energies on observable properties of chemical reaction networks, as given by microkinetics simulations based on ML-predicted reaction rates. After training an ANN to predict activation energies given standard molecular descriptors for reactants and products alone, we performed microkinetics simulations of three different prototypical reaction networks: formamide decomposition, aldol reactions and decomposition of 3-hydroperoxypropanal. We find that the kinetic modelling predictions can be in excellent agreement with corresponding simulations performed with ab initio calculations, but this is dependent on the inherent energetic landscape of the networks. We use these simulations to suggest some guidelines for when ML-based activation energies can be reliable, and when one should take more care in applications to kinetics modelling.

Enhanced-Sampling Simulations for the Estimation of Ligand Binding Kinetics: Current Status and Perspective

The dissociation rate (k off) associated with ligand unbinding events from proteins is a parameter of fundamental importance in drug design. Here we review recent major advancements in molecular simulation methodologies for the prediction of k off. Next, we discuss the impact of the potential energy function models on the accuracy of calculated k off values. Finally, we provide a perspective from high-performance computing and machine learning which might help improve such predictions.

A benchmark dataset for Hydrogen Combustion

The generation of reference data for deep learning models is challenging for reactive systems, and more so for combustion reactions due to the extreme conditions that create radical species and alternative spin states during the combustion process. Here, we extend intrinsic reaction coordinate (IRC) calculations with ab initio MD simulations and normal mode displacement calculations to more extensively cover the potential energy surface for 19 reaction channels for hydrogen combustion. A total of ∼290,000 potential energies and ∼1,270,000 nuclear force vectors are evaluated with a high quality range-separated hybrid density functional, ωB97X-V, to construct the reference data set, including transition state ensembles, for the deep learning models to study hydrogen combustion reaction.

Machine Learning of Reaction Properties via Learned Representations of the Condensed Graph of Reaction

The estimation of chemical reaction properties such as activation energies, rates, or yields is a central topic of computational chemistry. In contrast to molecular properties, where machine learning approaches such as graph convolutional neural networks (GCNNs) have excelled for a wide variety of tasks, no general and transferable adaptations of GCNNs for reactions have been developed yet. We therefore combined a popular cheminformatics reaction representation, the so-called condensed graph of reaction (CGR), with a recent GCNN architecture to arrive at a versatile, robust, and compact deep learning model. The CGR is a superposition of the reactant and product graphs of a chemical reaction and thus an ideal input for graph-based machine learning approaches. The model learns to create a data-driven, task-dependent reaction embedding that does not rely on expert knowledge, similar to current molecular GCNNs. Our approach outperforms current state-of-the-art models in accuracy, is applicable even to imbalanced reactions, and possesses excellent predictive capabilities for diverse target properties, such as activation energies, reaction enthalpies, rate constants, yields, or reaction classes. We furthermore curated a large set of atom-mapped reactions along with their target properties, which can serve as benchmark data sets for future work. All data sets and the developed reaction GCNN model are available online, free of charge, and open source.

Insight and Activation Energy Surface of the Dehydrogenation of C2HxO Species in Ethanol Oxidation Reaction on Ir(100)

Dehydrogenation of an organic compound is the first and the most fundamental elementary reaction in many organic reactions. In ethanol oxidation reaction (EOR) to form CO 2 , there are a total of 46 pathways in C 2 H x O (x=1-6) species leading to the removal of all six hydrogen atoms in five C-H bonds and one O-H bond. To investigate the degree of dehydrogenation in EOR under operando conditions, we performed density function theory (DFT) calculations to study 28 dehydrogenation steps of C 2 H x O on Ir(100). An activation energy surface was then constructed and compared with that of the C-C bond cleavages to understand the importance of the degree of dehydrogenation in EOR. The results show that there are likely 28 dehydrogenations in EOR under fuel cell temperatures and the last two hydrogens in C 2 H 2 O are less likely cleaved. On the other hand, deep dehydrogenation including 45 dehydrogenations can occur under ethanol steam reforming conditions.

Molecular Simulations to Elucidate Transport Phenomena in Polymeric Membranes

Despite decades of dominance in separation technology, progress in the design and development of high-performance polymer-based membranes has been incremental. Recent advances in materials science and chemical synthesis provide opportunities for molecular-level design of next-generation membrane materials. Such designs necessitate a fundamental understanding of transport and separation mechanisms at the molecular scale. Molecular simulations are important tools that could lead to the development of fundamental structure-property-performance relationships for advancing membrane design. In this Perspective, we assess the application and capability of molecular simulations to understand the mechanisms of ion and water transport across polymeric membranes. Additionally, we discuss the reliability of molecular models in mimicking the structure and chemistry of nanochannels and transport pathways in polymeric membranes. We conclude by providing research directions for resolving key knowledge gaps related to transport phenomena in polymeric membranes and for the construction of structure-property-performance relationships for the design of next-generation membranes.

Kernel charge equilibration: efficient and accurate prediction of molecular dipole moments with a machine-learning enhanced electron density model

State-of-the-art machine learning (ML) interatomic potentials use local representations of atomic environments to ensure linear scaling and size-extensivity. This implies a neglect of long-range interactions, most prominently related to electrostatics. To overcome this limitation, we herein present a ML framework for predicting charge distributions and their interactions termed kernel charge equilibration (kQEq). This model is based on classical charge equilibration (QEq) models expanded with an environment-dependent electronegativity. In contrast to previously reported neural network models with a similar concept, kQEq takes advantage of the linearity of both QEq and Kernel Ridge Regression to obtain a closed-form linear algebra expression for training the models. Furthermore, we avoid the ambiguity of charge partitioning schemes by using dipole moments as reference data. As a first application, we show that kQEq can be used to generate accurate and highly data-efficient models for molecular dipole moments.

Potential Application of Machine-Learning-Based Quantum Chemical Methods in Environmental Chemistry

It is an important topic in environmental sciences to understand the behavior and toxicology of chemical pollutants. Quantum chemical methodologies have served as useful tools for probing behavior and toxicology of chemical pollutants in recent decades. In recent years, machine learning (ML) techniques have brought revolutionary developments to the field of quantum chemistry, which may be beneficial for investigating environmental behavior and toxicology of chemical pollutants. However, the ML-based quantum chemical methods (ML-QCMs) have only scarcely been used in environmental chemical studies so far. To promote applications of the promising methods, this Perspective summarizes recent progress in the ML-QCMs and focuses on their potential applications in environmental chemical studies that could hardly be achieved by the conventional quantum chemical methods. Potential applications and challenges of the ML-QCMs in predicting degradation networks of chemical pollutants, searching global minima for atmospheric nanoclusters, discovering heterogeneous or photochemical transformation pathways of pollutants, as well as predicting environmentally relevant end points with wave functions as descriptors are introduced and discussed.

CARNOT: a Fragment-Based Direct Molecular Dynamics and Virtual-Reality Simulation Package for Reactive Systems

Traditionally, the study of reaction mechanisms of complex reaction systems such as combustion has been performed on an individual basis by optimizations of transition structure and minimum energy path or by reaction dynamics trajectory calculations for one elementary reaction at a time. It is effective, but time-consuming, whereas important and unexpected processes could have been missed. In this article, we present a direct molecular dynamics (DMD) approach and a virtual-reality simulation program, CARNOT, in which plausible chemical reactions are simulated simultaneously at finite temperature and pressure conditions. A key concept of the present ab initio molecular dynamics method is to partition a large, chemically reactive system into molecular fragments that can be adjusted on the fly of a DMD simulation. The theory represents an extension of the explicit polarization method to reactive events, called ReX-Pol. We propose a highest-and-lowest adapted-spin approximation to define the local spins of individual fragments, rather than treating the entire system by a delocalized wave function. Consequently, the present ab initio DMD can be applied to reactive systems consisting of an arbitrarily varying number of closed and open-shell fragments such as free radicals, zwitterions, and separate ions found in combustion and other reactions. A graph-data structure algorithm was incorporated in CARNOT for the analysis of reaction networks, suitable for reaction mechanism reduction. Employing the PW91 density functional theory and the 6-31+G(d) basis set, the capabilities of the CARNOT program were illustrated by a combustion reaction, consisting of 28 650 atoms, and by reaction network analysis that revealed a range of mechanistic and dynamical events. The method may be useful for applications to other types of complex reactions.

Improving machine learning performance on small chemical reaction data with unsupervised contrastive pretraining

Machine learning (ML) methods have great potential to transform chemical discovery by accelerating the exploration of chemical space and drawing scientific insights from data. However, modern chemical reaction ML models, such as those based on graph neural networks (GNNs), must be trained on a large amount of labelled data in order to avoid overfitting the data and thus possessing low accuracy and transferability. In this work, we propose a strategy to leverage unlabelled data to learn accurate ML models for small labelled chemical reaction data. We focus on an old and prominent problem-classifying reactions into distinct families-and build a GNN model for this task. We first pretrain the model on unlabelled reaction data using unsupervised contrastive learning and then fine-tune it on a small number of labelled reactions. The contrastive pretraining learns by making the representations of two augmented versions of a reaction similar to each other but distinct from other reactions. We propose chemically consistent reaction augmentation methods that protect the reaction center and find they are the key for the model to extract relevant information from unlabelled data to aid the reaction classification task. The transfer learned model outperforms a supervised model trained from scratch by a large margin. Further, it consistently performs better than models based on traditional rule-driven reaction fingerprints, which have long been the default choice for small datasets, as well as those based on reaction fingerprints derived from masked language modelling. In addition to reaction classification, the effectiveness of the strategy is tested on regression datasets; the learned GNN-based reaction fingerprints can also be used to navigate the chemical reaction space, which we demonstrate by querying for similar reactions. The strategy can be readily applied to other predictive reaction problems to uncover the power of unlabelled data for learning better models with a limited supply of labels.

Machine learning prediction of 3CLpro SARS-CoV-2 docking scores

Molecular docking results of two training sets containing 866 and 8,696 compounds were used to train three different machine learning (ML) approaches. Neural network approaches according to Keras and TensorFlow libraries and the gradient boosted decision trees approach of XGBoost were used with DScribe’s Smooth Overlap of Atomic Positions molecular descriptors. In addition, neural networks using the SchNetPack library and descriptors were used. The ML performance was tested on three different sets, including compounds for future organic synthesis. The final evaluation of the ML predicted docking scores was based on the ZINC in vivo set, from which 1,200 compounds were randomly selected with respect to their size. The results obtained showed a consistent ML prediction capability of docking scores, and even though compounds with more than 60 atoms were found slightly overestimated they remain valid for a subsequent evaluation of their drug repurposing suitability.

Autonomous Reaction Network Exploration in Homogeneous and Heterogeneous Catalysis

Autonomous computations that rely on automated reaction network elucidation algorithms may pave the way to make computational catalysis on a par with experimental research in the field. Several advantages of this approach are key to catalysis: (i) automation allows one to consider orders of magnitude more structures in a systematic and open-ended fashion than what would be accessible by manual inspection. Eventually, full resolution in terms of structural varieties and conformations as well as with respect to the type and number of potentially important elementary reaction steps (including decomposition reactions that determine turnover numbers) may be achieved. (ii) Fast electronic structure methods with uncertainty quantification warrant high efficiency and reliability in order to not only deliver results quickly, but also to allow for predictive work. (iii) A high degree of autonomy reduces the amount of manual human work, processing errors, and human bias. Although being inherently unbiased, it is still steerable with respect to specific regions of an emerging network and with respect to the addition of new reactant species. This allows for a high fidelity of the formalization of some catalytic process and for surprising in silico discoveries. In this work, we first review the state of the art in computational catalysis to embed autonomous explorations into the general field from which it draws its ingredients. We then elaborate on the specific conceptual issues that arise in the context of autonomous computational procedures, some of which we discuss at an example catalytic system.

GRAPHICAL ABSTRACT

SUPPLEMENTARY INFORMATION

The online version contains supplementary material available at 10.1007/s11244-021-01543-9.

Prospects and challenges for autonomous catalyst discovery viewed from an experimental perspective

Autonomous catalysis research requires elaborate integration of operando experiments into automated workflows. Suitable experimental data for analysis by artificial intelligence can be measured more readily according to standard operating procedures.

Data-efficient iterative training of Gaussian approximation potentials: Application to surface structure determination of rutile IrO2 and RuO2

Machine-learning interatomic potentials, such as Gaussian Approximation Potentials (GAPs), constitute a powerful class of surrogate models to computationally involved first-principles calculations. At a similar predictive quality but significantly reduced cost, they could leverage otherwise barely tractable extensive sampling as in global surface structure determination (SSD). This efficiency is jeopardized though, if an a priori unknown structural and chemical search space as in SSD requires an excessive number of first-principles data for the GAP training. To this end, we present a general and data-efficient iterative training protocol that blends the creation of new training data with the actual surface exploration process. Demonstrating this protocol with the SSD of low-index facets of rutile IrO₂ and RuO₂, the involved simulated annealing on the basis of the refining GAP identifies a number of unknown terminations even in the restricted sub-space of (1 × 1) surface unit cells. Particularly in an O-poor environment, some of these, then metal-rich terminations, are thermodynamically most stable and are reminiscent of complexions as discussed for complex ceramic materials.

Progress towards machine learning reaction rate constants

Quantum and classical reaction rate constant calculations come at the cost of exploring potential energy surfaces. Due to the "curse of dimensionality", their evaluation quickly becomes unfeasible as the system size grows. Machine learning algorithms can accelerate the calculation of reaction rate constants by predicting them using low cost input features. In this perspective, we briefly introduce supervised machine learning algorithms in the context of reaction rate constant prediction. We discuss existing and recently created kinetic datasets and input feature representations as well as the use and design of machine learning algorithms to predict reaction rate constants or quantities required for their computation. Amongst these, we first describe the use of machine learning to predict activation, reaction, solvation and dissociation energies. We then look at the use of machine learning to predict reactive force field parameters, reaction rate constants as well as to help accelerate the search for minimum energy paths. Lastly, we provide an outlook on areas which have yet to be explored so as to improve and evaluate the use of machine learning algorithms for chemical reaction rate constants.

Toward Precise n-Type Doping Control in MOVPE-Grown β-Ga2O3 Thin Films by Deep-Learning Approach

In this work, we train a hybrid deep-learning model (fDNN, Forest Deep Neural Network) to predict the doping level measured from the Hall Effect measurement at room temperature and to investigate the doping behavior of Si dopant in both (100) and (010) β-Ga2O3 thin film grown by the metalorganic vapor phase epitaxy (MOVPE). The model reveals that a hidden parameter, the Si supplied per nm (mol/nm), has a dominant influence on the doping process compared with other process parameters. An empirical relation is concluded from this model to estimate the doping level of the grown film with the Si supplied per nm (mol/nm) as the primary variable for both (100) and (010) β-Ga2O3 thin film. The outcome of the work indicates the similarity between the doping behavior of (100) and (010) β-Ga2O3 thin film via MOVPE and the generality of the results to different deposition systems.