Reaction Templates: Bridging Synthesis Knowledge and Artificial Intelligence.

ConspectusThe field of chemical research boasts a long history of developing software to automate synthesis planning and reaction prediction. Early software relied heavily on expert systems, requiring significant effort to encode vast amounts of synthesis knowledge into a computer-readable format. However, recent advancements in deep learning have shifted the focus toward AI models, offering improved prediction capabilities. Despite these advancements, current AI models often lack the integration of known synthesis rules and intuitions, creating a gap that hinders interpretability and future development of the models. To bridge them, our research group has been actively working on incorporating reaction templates into deep learning models, achieving promising results across various applications.In this Account, we present our latest works to incorporate the known synthesis knowledge into the deep learning models through the utilization of reaction templates. We begin by highlighting the limitations of early computer programs heavily reliant on hand-coded rules. These programs, while providing a foundation for the field, presented limitations in scalability and adaptability. We then introduce SMARTS (SMILES arbitrary target specification), a popular Python-readable format for representing chemical reactions. This format of reaction encoding facilitates the quick integration of synthesis knowledge into AI models built using the Python language. With the SMARTS-based reaction templates, we introduce our recent efforts of developing an AI model for reaction-based molecule optimization. Subsequently, we discuss the recent efforts to automate the extraction of reaction templates from vast chemical reaction databases. This approach eliminates the previously required manual effort of encoding knowledge, a process that could be time-consuming and prone to error when dealing with large data sets. By customizing the automated extraction algorithm, we have developed powerful AI models for specific tasks such as retrosynthesis (LocalRetro), reaction outcome prediction (LocalTransform), and atom-to-atom mapping (LocalMapper). These models, aligned with the intuition of chemists, demonstrate the effectiveness of incorporating reaction templates into deep learning frameworks.Looking toward the future, we believe that utilizing reaction templates to connect known chemical knowledge and AI models holds immense potential for various applications. Not only can this approach significantly benefit future AI models focused on challenging tasks like reaction mechanism labeling and prediction, but we anticipate it can also extend its reach to the realm of inorganic synthesis. By integrating synthesis knowledge, we can not only achieve improved performance but also enhance the interpretability of AI models, paving the way for further advancements in AI-powered chemical synthesis.

Data-Driven Prediction of Configurational Stability of Molecule-Adsorbed Heterogeneous Catalysts.

The design of new heterogeneous catalysts that convert small molecules into valuable chemicals is a key challenge for constructing sustainable energy systems. Density functional theory (DFT)-based design frameworks based on the understanding of molecular adsorption on the catalytic surface have been widely proposed to accelerate experimental approaches to develop novel catalysts. In addition, a machine learning (ML)-combined design framework was recently proposed to further reduce the inherent time cost of DFT-based frameworks. However, because of the lack of prior information on chemical interactions between arbitrary surfaces and adsorbates, the efficacy of the computational screening approaches would be reduced by obtaining unexpected structural anomalies (i.e., abnormally converged surface-adsorbate geometries after the DFT calculations) during an exhaustive exploration of chemical space. To overcome this challenge, we propose an ML framework that directly predicts the configurational stability of a given initial surface-adsorbate geometry. Our benchmark experiments with the Open Catalysts 20 (OC20) dataset show promising performance on classifying stable geometry (i.e., F1-score of 0.922, the area under the receiver operating characteristics (AUROC) of 0.906, and Matthews correlation coefficient (MCC) of 0.633) with a high precision of 0.921 by utilizing an ensemble approach. We further interpret the generalizability and domain applicability of the trained model in terms of the chemical space of the OC20 dataset. Furthermore, from an experiment on the training set size dependence of model performance, we found that our ML model could be practically applicable to classify stable configurations even with a relatively small number of training data.

Autobifunctional Mechanism of Jagged Pt Nanowires for Hydrogen Evolution Kinetics via End-to-End Simulation.

The extraordinary mass activity of jagged Pt nanowires can substantially improve the economics of the hydrogen evolution reaction (HER). However, it is a great challenge to fully unveil the HER kinetics driven by the jagged Pt nanowires with their multiscale morphology. Herein we present an end-to-end framework that combines experiment, machine learning, and multiscale advances of the past decade to elucidate the HER kinetics catalyzed by jagged Pt nanowires under alkaline conditions. The bifunctional catalysis conventionally refers to the synergistic increase in activity by the combination of two different catalysts. We report that monometals, such as jagged Pt nanowires, can exhibit bifunctional characteristics owing to its complex surface morphology, where one site prefers electrochemical proton adsorption and another is responsible for activation, resulting in a 4-fold increase in the activity. We find that the conventional design guideline that the sites with a 0 eV Gibbs free energy of adsorption are optimal for HER does not hold under alkaline conditions, and rather, an energy between -0.2 and 0.0 eV is shown to be optimal. At the reaction temperatures, the high activity arises from low-coordination-number (≤7) Pt atoms exposed by the jagged surface. Our current demonstration raises an emerging prospect to understand highly complex kinetic phenomena on the nanoscale in full by implementing end-to-end multiscale strategies.

Trends of age-related non-communicable diseases in people living with HIV and comparison with uninfected controls: A nationwide population-based study in South Korea.

OBJECTIVES: We aim to compare the trends of non-communicable diseases (NCDs) and death among people living with HIV (PLWH) and uninfected controls in South Korea. METHODS: We identified PLWH from a nationwide database of all Korean citizens enrolled from 1 January 2004 to 31 December 2016. A control cohort was randomly selected for PLWH by frequency matching for age and sex in a 20:1 ratio. To compare NCD trends between the groups, adjusted incidence rate ratios for outcomes across ages, calendar years and times after HIV diagnosis were calculated. RESULTS: We included 14 134 PLWH and 282 039 controls in this study; 58.5% of PLWH and 36.4% of the controls were diagnosed with at least one NCD. The incidence rates of cancers, chronic kidney disease, depression, osteoporosis, diabetes and dyslipidaemia were higher in PLWH than in the controls, whereas those of cardiovascular disease, heart failure, ischaemic stroke and hypertension were lower in PLWH. Relative risks (RRs) for NCDs in PLWH were higher than controls in younger age groups. Trends in the RRs of NCDs tended to increase with the calendar year for PLWH vs. controls and either stabilized or decreased with time after HIV diagnosis. The RR of death from PLWH has decreased with the calendar year, but showed a tendency to rise again after 2014 and was significant at the early stage of HIV diagnosis. CONCLUSIONS: Although the RR of each NCD in PLWH showed variable trends compared with that in controls, NCDs in PLWH have been increasingly prevalent.

Structure-Based Synthesizability Prediction of Crystals Using Partially Supervised Learning.

Predicting the synthesizability of inorganic materials is one of the major challenges in accelerated material discovery. A widely employed approximate approach is to consider the thermodynamic decomposition stability due to its simplicity of computing, but it is notorious for either producing too many candidates or missing important metastable materials. These results, however, are not unexcepted since the synthesizability is a complex phenomenon, and the thermodynamic stability is just one contributor. Here, we suggest a machine-learning model to quantify the probability of synthesis based on the partially supervised learning of materials database. We adapted the positive and unlabeled machine learning (PU learning) by implementing the graph convolutional neural network as a classifier in which the model outputs crystal-likeness scores (CLscore). The model shows 87.4% true positive (CLscore > 0.5) prediction accuracy for the test set of experimentally reported cases (9356 materials) in the Materials Project. We further validated the model by predicting the synthesizability of newly reported experimental materials in the last 5 years (2015-2019) with an 86.2% true positive rate using the model trained with the database as of the end of year 2014. Our analysis shows that our model captures the structural motif for synthesizability beyond what is possible by Ehull. We find that 71 materials among the top 100 high-scoring virtual materials have indeed been previously synthesized in the literature. With the proposed data-driven metric of the crystal-likeness score, high-throughput virtual screenings and generative models can benefit significantly by effectively reducing the chemical space that needs to be explored experimentally in the future toward more rational materials design.

Uncertainty-Quantified Hybrid Machine Learning/Density Functional Theory High Throughput Screening Method for Crystals.

Computational high throughput screening (HTS) has emerged as a significant tool in material science to accelerate the discovery of new materials with target properties in recent years. However, despite many successful cases in which HTS led to the novel discovery, currently, the major bottleneck in HTS is a large computational cost of density functional theory (DFT) calculations that scale cubically with system size, limiting the chemical space that can be explored. The present work aims at addressing this computational burden of HTS by presenting a machine learning (ML) framework that can efficiently explore the chemical space. Our model is built upon an existing crystal graph convolutional neural network (CGCNN) to obtain formation energy of a crystal structure but is modified to allow uncertainty quantification for each prediction using the hyperbolic tangent activation function and dropout algorithm (CGCNN-HD). The uncertainty quantification is particularly important since typical usage of CGCNN (due to the lack of gradient implementation) does not involve structural relaxation which could cause substantial prediction errors. The proposed method is benchmarked against an existing application that identified promising photoanode material among the >7,000 hypothetical Mg-Mn-O ternary compounds using all DFT-HTS. In our approach, we perform the approximate HTS using CGCNN-HD and refine the results using full DFT for those selected (denoted as ML/DFT-HTS). The proposed hybrid model reduces the required DFT calculations by a factor of >50 compared to the previous DFT-HTS in making the same discovery of Mg2MnO4, experimentally validated new photoanode material. Further analysis demonstrates that the addition of HD components with uncertainty measures in the CGCNN-HD model increased the discoverability of promising materials relative to all DFT-HTS from 30% (CGCNN) to 68% (CGCNN-HD). The present ML/DFT-HTS with uncertainty quantification can thus be a fast alternative to DFT-HTS for efficient exploration of the vast chemical space.

Long-term Effects of Cumulative Average PM2.5 Exposure on the Risk of Hemorrhagic Stroke.

BACKGROUND: Epidemiological studies have revealed associations between the fine particle (PM2.5; aerodynamic diameter <2.5 µm) exposure and cardiovascular disease. Researchers have also recently begun investigating the association between PM2.5 exposure and hemorrhagic stroke (HS) and identifying subpopulations vulnerable to PM2.5 exposure. Long-term cumulative average PM2.5 exposure may affect the risk of HS, and these effects may be modified by risk factors. METHODS: This retrospective study evaluated the effects of PM2.5 on the time-to-first-diagnosis of HS among 62,676 Seoul metropolitan city residents with 670,431 total person-years of follow-up; this cohort is a subset from a nationally representative cohort of 1,025,340 individuals from the Korean National Health Insurance Service database (2002-2013). A time-dependent Cox proportional hazards model was used to adjust for age, sex, household income, insurance type, body mass index, smoking status, medical history, and family history. The annual mean PM2.5 concentrations for 25 districts were used as the time-dependent variable. Subgroup analyses of the traditional risk factors of HS were performed to evaluate potential effect modifications. RESULTS: Each 10-µg/m increment in cumulative average PM2.5 exposure was noticeably associated with HS (hazard ratio [HR] = 1.43; 95% confidence interval [CI]: 1.09-1.88). The adverse effects of PM2.5 exposure were modified by ≥65 years of age (HR = 2.00; 95% CI = 1.32, 3.02) and obesity (body mass index ≥25 kg/m; HR = 1.91; 95% CI = 1.28, 2.84). CONCLUSIONS: Cumulative average PM2.5 exposure might increase the risk of HS. Elderly (≥65 years) and obese individuals may be more vulnerable to the effects of PM2.5 exposure.

High and low sodium intakes are associated with incident chronic kidney disease in patients with normal renal function and hypertension.

The association between salt intake and renal outcome in subjects with preserved kidney function remains unclear. Here we evaluated the effect of sodium intake on the development of chronic kidney disease (CKD) in a prospective cohort of people with normal renal function. Data were obtained from the Korean Genome and Epidemiology Study, a prospective community-based cohort study while sodium intake was estimated by a 24-hour dietary recall Food Frequency Questionnaire. A total of 3,106 individuals with and 4,871 patients without hypertension were analyzed with a primary end point of CKD development [a composite of estimated glomerular filtration rate (eGFR) under 60 mL/min/1.73 m2 and/or development of proteinuria during follow-up]. The median ages were 55 and 47 years, the proportions of males 50.9% and 46.3%, and the median eGFR 92 and 96 mL/min/1.73 m2 in individuals with and without hypertension, respectively. During a median follow-up of 123 months in individuals with hypertension and 140 months in those without hypertension, CKD developed in 27.8% and 16.5%, respectively. After adjusting for confounders, multiple Cox models indicated that the risk of CKD development was significantly higher in people with hypertension who consumed less than 2.08 g/day or over 4.03 g/day sodium than in those who consumed between 2.93-4.03 g/day sodium. However, there was no significant difference in the incident CKD risk among each quartile of people without hypertension. Thus, both high and low sodium intakes were associated with increased risk for CKD, but this relationship was only observed in people with hypertension.

Warfarin Use in Patients With Atrial Fibrillation Undergoing Hemodialysis: A Nationwide Population-Based Study.

BACKGROUND AND PURPOSE: The aim of this study is to elucidate the effects of warfarin use in patients with atrial fibrillation undergoing dialysis using a population-based Korean registry. METHODS: Data were extracted from the Health Insurance Review and Assessment Service, which is a nationwide, mandatory social insurance database of all Korean citizens enrolled in the National Health Information Service between 2009 and 2013. Thromboembolic and hemorrhagic outcomes were analyzed according to warfarin use. Overall and propensity score-matched cohorts were analyzed by Cox proportional hazards models. RESULTS: Among 9974 hemodialysis patients with atrial fibrillation, the mean age was 66.6±12.2 years, 5806 (58.2%) were men, and 2921 (29.3%) used warfarin. After propensity score matching to adjust for all described baseline differences, 5548 subjects remained, and differences in baseline variables were distributed equally between warfarin users and nonusers. During a mean follow-up duration of 15.9±11.1 months, ischemic and hemorrhagic stroke occurred in 678 (6.8%) and 227 (2.3%) patients, respectively. In a multiple Cox model, warfarin use was significantly associated with an increased risk of hemorrhagic stroke (hazard ratio, 1.44; 95% confidence interval, 1.09-1.91; P=0.010) in the overall cohort. Furthermore, a significant relationship between warfarin use and hemorrhagic stroke was found in propensity-matched subjects (hazard ratio, 1.56; 95% confidence interval, 1.10-2.22; P=0.013). However, the ratios for ischemic stroke were not significantly different in either the propensity-matched (hazard ratio, 0.95; 95% confidence interval, 0.78-1.15; P=0.569) or overall cohort (hazard ratio, 1.06; 95% confidence interval, 0.90-1.26; P=0.470). CONCLUSIONS: Our findings suggest that warfarin should be used carefully in hemodialysis patients, given the higher risk of hemorrhagic events and the lack of ability to prevent thromboembolic complications.

Predicting synthesis recipes of inorganic crystal materials using elementwise template formulation.

While advances in computational techniques have accelerated virtual materials design, the actual synthesis of predicted candidate materials is still an expensive and slow process. While a few initial studies attempted to predict the synthesis routes for inorganic crystals, the existing models do not yield the priority of predictions and could produce thermodynamically unrealistic precursor chemicals. Here, we propose an element-wise graph neural network to predict inorganic synthesis recipes. The trained model outperforms the popularity-based statistical baseline model for the top-k exact match accuracy test, showing the validity of our approach for inorganic solid-state synthesis. We further validate our model by the publication-year-split test, where the model trained based on the materials data until the year 2016 is shown to successfully predict synthetic precursors for the materials synthesized after 2016. The high correlation between the probability score and prediction accuracy suggests that the probability score can be interpreted as a measure of confidence levels, which can offer the priority of the predictions.

Accelerated chemical science with AI.

In light of the pressing need for practical materials and molecular solutions to renewable energy and health problems, to name just two examples, one wonders how to accelerate research and development in the chemical sciences, so as to address the time it takes to bring materials from initial discovery to commercialization. Artificial intelligence (AI)-based techniques, in particular, are having a transformative and accelerating impact on many if not most, technological domains. To shed light on these questions, the authors and participants gathered in person for the ASLLA Symposium on the theme of 'Accelerated Chemical Science with AI' at Gangneung, Republic of Korea. We present the findings, ideas, comments, and often contentious opinions expressed during four panel discussions related to the respective general topics: 'Data', 'New applications', 'Machine learning algorithms', and 'Education'. All discussions were recorded, transcribed into text using Open AI's Whisper, and summarized using LG AI Research's EXAONE LLM, followed by revision by all authors. For the broader benefit of current researchers, educators in higher education, and academic bodies such as associations, publishers, librarians, and companies, we provide chemistry-specific recommendations and summarize the resulting conclusions.

Correction to "Generative Adversarial Networks for Crystal Structure Prediction".

[This corrects the article DOI: 10.1021/acscentsci.0c00426.].

Bimetallic Gold-Silver Nanostructures Drive Low Overpotentials for Electrochemical Carbon Dioxide Reduction.

Alloy formation is an advanced approach to improve desired properties that the monoelements cannot achieve. Alloys are usually designed to tailor intrinsic natures or induce synergistic effects by combining materials with distinct properties. Indeed, unprecedented properties have emerged in many cases, superior to a simple sum of pure elements. Here, we present Au-Ag alloy nanostructures with prominent catalytic properties in an electrochemical carbon dioxide reduction reaction (eCO2RR). The Au-Ag hollow nanocubes are prepared by galvanic replacement of Au on Ag nanocubes. When the Au-to-Ag ratio is 1:1 (Au1Ag1), the alloy hollow nanocubes exhibit maximum Faradaic efficiencies of CO production in a wide potential range and high mass activity and CO current density superior to those of the bare metals. In particular, overpotentials are estimated to be similar to or lower than that of the Au catalyst under various standard metrics. Density functional theory calculations, machine learning, and a statistical consideration demonstrate that the optimal configuration of the *COOH intermediate is a bidentate coordination structure where C binds to Au and O binds to Ag. This active Au-Ag neighboring configuration has a maximum population and enhanced intrinsic catalytic activity on the Au1Ag1 surface among other Au-to-Ag compositions, in good agreement with the experimental results. Further application of Au1Ag1 to a membrane electrode assembly cell at neutral conditions shows enhanced CO Faradaic efficiency and current densities compared to Au or Ag nanocubes, indicating the possible extension of Au-Ag alloys to larger electrochemical systems. These results give a new insight into the synergistic roles of Au and Ag in the eCO2RR and offer a fresh direction toward a rational design of bimetallic catalysts at a practical scale.

Optimal treatment recommendations for diabetes patients using the Markov decision process along with the South Korean electronic health records.

The extensive utilization of electronic health records (EHRs) and the growth of enormous open biomedical datasets has readied the area for applications of computational and machine learning techniques to reveal fundamental patterns. This study's goal is to develop a medical treatment recommendation system using Korean EHRs along with the Markov decision process (MDP). The sharing of EHRs by the National Health Insurance Sharing Service (NHISS) of Korea has made it possible to analyze Koreans' medical data which include treatments, prescriptions, and medical check-up. After considering the merits and effectiveness of such data, we analyzed patients' medical information and recommended optimal pharmaceutical prescriptions for diabetes, which is known to be the most burdensome disease for Koreans. We also proposed an MDP-based treatment recommendation system for diabetic patients to help doctors when prescribing diabetes medications. To build the model, we used the 11-year Korean NHISS database. To overcome the challenge of designing an MDP model, we carefully designed the states, actions, reward functions, and transition probability matrices, which were chosen to balance the tradeoffs between reality and the curse of dimensionality issues.

Understanding potential-dependent competition between electrocatalytic dinitrogen and proton reduction reactions.

A key challenge to realizing practical electrochemical N2 reduction reaction (NRR) is the decrease in the NRR activity before reaching the mass-transfer limit as overpotential increases. While the hydrogen evolution reaction (HER) has been suggested to be responsible for this phenomenon, the mechanistic origin has not been clearly explained. Herein, we investigate the potential-dependent competition between NRR and HER using the constant electrode potential model and microkinetic modeling. We find that the H coverage and N2 coverage crossover leads to the premature decrease of NRR activity. The coverage crossover originates from the larger charge transfer in H+ adsorption than N2 adsorption. The larger charge transfer in H+ adsorption, which potentially leads to the coverage crossover, is a general phenomenon seen in various heterogeneous catalysts, posing a fundamental challenge to realize practical electrochemical NRR. We suggest several strategies to overcome the challenge based on the present understandings.

Predicting potentially hazardous chemical reactions using an explainable neural network.

Predicting potentially dangerous chemical reactions is a critical task for laboratory safety. However, a traditional experimental investigation of reaction conditions for possible hazardous or explosive byproducts entails substantial time and cost, for which machine learning prediction could accelerate the process and help detailed experimental investigations. Several machine learning models have been developed which allow the prediction of major chemical reaction products with reasonable accuracy. However, these methods may not present sufficiently high accuracy for the prediction of hazardous products which particularly requires a low false negative result for laboratory safety in order not to miss any dangerous reactions. In this work, we propose an explainable artificial intelligence model that can predict the formation of hazardous reaction products in a binary classification fashion. The reactant molecules are transformed into substructure-encoded fingerprints and then fed into a convolutional neural network to make the binary decision of the chemical reaction. The proposed model shows a false negative rate of 0.09, which can be compared with 0.47-0.66 using the existing main product prediction models. To provide explanations for what substructures of the given reactant molecules are important to make a decision for target hazardous product formation, we apply an input attribution method, layer-wise relevance propagation, which computes the contributions of individual inputs per input data. The computed attributions indeed match some of the existing chemical intuitions and mechanisms, and also offer a way to analyze possible data-imbalance issues of the current predictions based on relatively small positive datasets. We expect that the proposed hazardous product prediction model will be complementary to existing main product prediction models and experimental investigations.

Mortality and morbidity in children with asthma: A nationwide study in Korea.

OBJECTIVE: Childhood mortality due to asthma remains problematic; however, asthma-related mortality in Korean children has not been previously described. This study aimed to estimate asthma mortality and morbidity and determine to what extent asthma contributes to childhood mortality in Korea. METHODS: Data from 9 to 12 million children (aged < 18 years) per year recorded for each year between 2002 and 2015 were retrieved from the Korea National Health Insurance claims database. Patients with asthma during the year preceding death were investigated. Causes of death were analysed using data obtained from the Korean Statistical Information Service database. Cause-specific mortality was examined, and the mortality rate of children with asthma was compared to that of the general paediatric population with respect to the cause of death and age. Hospital use by patients with asthma, including intensive care unit admission and hospitalisation, was analysed. RESULTS: Asthma mortality decreased from 0.09 per 100,000 children in 2003 to 0.02 per 100,000 children in 2014, with an average mortality of 0.06 per 100,000 children. Mortality due to respiratory diseases was four times more common in patients with asthma than in the general population of children aged >5 years, despite decreases in asthma-related mortality. Asthma-related hospitalisations and intensive care interventions tended to decrease throughout the study period. CONCLUSIONS: Asthma mortality declined in children between 2003 and 2015 in Korea. Children with asthma are at a higher risk of death from respiratory diseases than the general population.

Cohort profile: the Environmental-Pollution-Induced Neurological EFfects (EPINEF) study: a multicenter cohort study of Korean adults.

The general population is exposed to numerous environmental pollutants, and it remains unclear which pollutants affect the brain, accelerating brain aging and increasing the risk of dementia. The Environmental-Pollution-Induced Neurological Effects study is a multi-city prospective cohort study aiming to comprehensively investigate the effect of different environmental pollutants on brain structures, neuropsychological function, and the development of dementia in adults. The baseline data of 3,775 healthy elderly people were collected from August 2014 to March 2018. The eligibility criteria were age ≥50 years and no self-reported history of dementia, movement disorders, or stroke. The assessment included demographics and anthropometrics, laboratory test results, and individual levels of exposure to air pollution. A neuroimaging sub-cohort was also recruited with 1,022 participants during the same period, and brain magnetic resonance imaging and neuropsychological tests were conducted. The first follow-up environmental pollutant measurements will start in 2022 and the follow-up for the sub-cohort will be conducted every 3-4 years. We have found that subtle structural changes in the brain may be induced by exposure to airborne pollutants such as particulate matter 10 µm or less in diameter (PM10), particulate matter 2.5 µm or less in diameter (PM2.5) and Mn10, manganese in PM10; Mn2.5, manganese in PM2.5. PM10, PM2.5, and nitrogen dioxide in healthy adults. This study provides a basis for research involving large-scale, long-term neuroimaging assessments in community-based populations.

Machine-enabled inverse design of inorganic solid materials: promises and challenges.

Developing high-performance advanced materials requires a deeper insight and search into the chemical space. Until recently, exploration of materials space using chemical intuitions built upon existing materials has been the general strategy, but this direct design approach is often time and resource consuming and poses a significant bottleneck to solve the materials challenges of future sustainability in a timely manner. To accelerate this conventional design process, inverse design, which outputs materials with pre-defined target properties, has emerged as a significant materials informatics platform in recent years by leveraging hidden knowledge obtained from materials data. Here, we summarize the latest progress in machine-enabled inverse materials design categorized into three strategies: high-throughput virtual screening, global optimization, and generative models. We analyze challenges for each approach and discuss gaps to be bridged for further accelerated and rational data-driven materials design.