ChatMOF: an artificial intelligence system for predicting and generating metal-organic frameworks using large language models.

ChatMOF is an artificial intelligence (AI) system that is built to predict and generate metal-organic frameworks (MOFs). By leveraging a large-scale language model (GPT-4, GPT-3.5-turbo, and GPT-3.5-turbo-16k), ChatMOF extracts key details from textual inputs and delivers appropriate responses, thus eliminating the necessity for rigid and formal structured queries. The system is comprised of three core components (i.e., an agent, a toolkit, and an evaluator) and it forms a robust pipeline that manages a variety of tasks, including data retrieval, property prediction, and structure generations. ChatMOF shows high accuracy rates of 96.9% for searching, 95.7% for predicting, and 87.5% for generating tasks with GPT-4. Additionally, it successfully creates materials with user-desired properties from natural language. The study further explores the merits and constraints of utilizing large language models (LLMs) in combination with database and machine learning in material sciences and showcases its transformative potential for future advancements.

RNA-Seq of an LPS-Induced Inflammation Model Reveals Transcriptional Profile Patterns of Inflammatory Processes.

The LPS-induced inflammation model is widely used for studying inflammatory processes due to its cost-effectiveness, reproducibility, and faithful representation of key hallmarks. While researchers often validate this model using clinical cytokine markers, a comprehensive understanding of gene regulatory mechanisms requires extending investigation beyond these hallmarks. Our study leveraged multiple whole-blood bulk RNA-seq datasets to rigorously compare the transcriptional profiles of the well-established LPS-induced inflammation model with those of several human diseases characterized by systemic inflammation. Beyond conventional inflammation-associated systems, we explored additional systems indirectly associated with inflammatory responses (i.e., ISR, RAAS, and UPR) using a customized core inflammatory gene list. Our cross-condition-validation approach spanned four distinct conditions: systemic lupus erythematosus (SLE) patients, dengue infection, candidemia infection, and staphylococcus aureus exposure. This analysis approach, utilizing the core gene list aimed to assess the model's suitability for understanding the gene regulatory mechanisms underlying inflammatory processes triggered by diverse factors. Our analysis resulted in elevated expressions of innate immune-associated genes, coinciding with suppressed expressions of adaptive immune-associated genes. Also, upregulation of genes associated with cellular stresses and mitochondrial innate immune responses underscored oxidative stress as a central driver of the corresponding inflammatory processes in both the LPS-induced and other inflammatory contexts.

Multi-System-Level Analysis Reveals Differential Expression of Stress Response-Associated Genes in Inflammatory Solar Lentigo.

Although the pathogenesis of solar lentigo (SL) involves chronic ultraviolet (UV) exposure, cellular senescence, and upregulated melanogenesis, underlying molecular-level mechanisms associated with SL remain unclear. The aim of this study was to investigate the gene regulatory mechanisms intimately linked to inflammation in SL. Skin samples from patients with SL with or without histological inflammatory features were obtained. RNA-seq data from the samples were analyzed via multiple analysis approaches, including exploration of core inflammatory gene alterations, identifying functional pathways at both transcription and protein levels, comparison of inflammatory module (gene clusters) activation levels, and analyzing correlations between modules. These analyses disclosed specific core genes implicated in oxidative stress, especially the upregulation of nuclear factor kappa B in the inflammatory SLs, while genes associated with protective mechanisms, such as SLC6A9, were highly expressed in the non-inflammatory SLs. For inflammatory modules, Extracellular Immunity and Mitochondrial Innate Immunity were exclusively upregulated in the inflammatory SL. Analysis of protein-protein interactions revealed the significance of CXCR3 upregulation in the pathogenesis of inflammatory SL. In conclusion, the upregulation of stress response-associated genes and inflammatory pathways in response to UV-induced oxidative stress implies their involvement in the pathogenesis of inflammatory SL.

Multimodal Transformer for Property Prediction in Polymers.

In this work, we designed a multimodal transformer that combines both the Simplified Molecular Input Line Entry System (SMILES) and molecular graph representations to enhance the prediction of polymer properties. Three models with different embeddings (SMILES, SMILES + monomer, and SMILES + dimer) were employed to assess the performance of incorporating multimodal features into transformer architectures. Fine-tuning results across five properties (i.e., density, glass-transition temperature (Tg), melting temperature (Tm), volume resistivity, and conductivity) demonstrated that the multimodal transformer with both the SMILES and the dimer configuration as inputs outperformed the transformer using only SMILES across all five properties. Furthermore, our model facilitates in-depth analysis by examining attention scores, providing deeper insights into the relationship between the deep learning model and the polymer attributes. We believe that our work, shedding light on the potential of multimodal transformers in predicting polymer properties, paves a new direction for understanding and refining polymer properties.

Dual-Photosensitizer Synergy Empowers Ambient Light Photoactivation of Indium Oxide for High-Performance NO2 Sensing.

Light-activated chemiresistors offer a powerful approach to achieving lower-temperature gas sensing with unprecedented sensitivities. However, an incomplete understanding of how photoexcited charge carriers enhance sensitivity obstructs the rational design of high-performance sensors, impeding the practical utilization under commonly accessible light sources instead of ultraviolet or higher-energy sources. Here, a rational approach is presented to modulate the electronic properties of the parent metal oxide phase, exemplified by this model system of Bi-doped In2 O3 nanofibers decorated with Au nanoparticles (NPs) that exhibit superior NO2 sensing performance. Bi doping introduces mid-gap energy levels into In2 O3 , promoting photoactivation even under visible blue light. Additionally, green-absorbing plasmonic Au NPs facilitate electron transfer across the heterojunction, extending the photoactive region toward the green light. It is revealed that the direct involvement of photogenerated charge carriers in gas adsorption and desorption processes is pivotal for enhancing gas sensing performance. Owing to the synergistic interplay between the Bi dopants and the Au NPs, the Au-Bix In2-x O3 (x = 0.04) sensing layers attain impressive response values (Rg /Ra  = 104 at 0.6 ppm NO2 ) under green light illumination and demonstrate practical viability through evaluation under simulated mixed-light conditions, all of which significantly outperforms previously reported visible light-activated NO2 sensors.

Unveiling interfacial interaction between antimony oxyanions and boehmite nanorods: Spectroscopic evidence and density functional theory analysis.

In natural environments, the fate and migratory behavior of metalloid contaminants such as antimony (Sb) significantly depend on the interfacial reactivity of mineral surfaces. Although boehmite (γ-AlOOH) is widely observed in (sub)surface environments, its underlying interaction mechanism with Sb oxyanions at the molecular scale remains unclear. Considering Sb-contaminated environmental conditions in this study, we prepared boehmite under weakly acidic conditions for use in the systematic investigation of interfacial interactions with Sb(III) and Sb(V). The as-synthesized boehmite showed a nanorod morphology and comprised four crystal facets in the following order: 48.4% (010), 27.1% (101), 15.0% (001), and 9.5% (100). The combined results of spectroscopic analyses and theoretical calculations revealed that Sb(III) formed hydrogen bonding outer-sphere complexation on the (100), (010), and (001) facets and that Sb(V) preferred to form bidentate inner-sphere complexation via mononuclear edge-sharing configuration on the (100), (001), and (101) facets and binuclear corner-sharing configuration on the (010) facet. These findings indicate that the facet-mediated thermodynamic stability of the surface complexation determines the interaction affinity toward the Sb species. This work is the first to document the contribution of boehmite to (sub)surface media, improving the ability to forecast the fate and behavior of Sb oxyanions at mineral-water interfaces.

Structural Determinants of Chirally Selective Transport of Amino Acids through the α-Hemolysin Protein Nanopores of Free-Standing Planar Lipid Membranes.

Despite the importance of the enantioselective transport of amino acids through transmembrane protein nanopores from fundamental and practical perspectives, little has been explored to date. Here, we study the transport of amino acids through α-hemolysin (αHL) protein pores incorporated into a free-standing lipid membrane. By measuring the transport of 13 different amino acids through the αHL pores, we discover that the molecular size of the amino acids and their capability to form hydrogen bonds with the pore surface determine the chiral selectivity. Molecular dynamics simulations corroborate our findings by revealing the enantioselective molecular-level interactions between the amino acid enantiomers and the αHL pore. Our work is the first to present the determinants for chiral selectivity using αHL protein as a molecular filter.

Boosting Alkaline Hydrogen Oxidation Activity of Ru Single-Atom Through Promoting Hydroxyl Adsorption on Ru/WC1- x Interfaces.

The sluggish kinetics of the hydrogen oxidation reaction (HOR) in alkaline conditions continue to pose a significant challenge for the practical implementation of anion-exchange membrane fuel cells. Developing single-atom catalysts can accelerate the pace of new HOR catalyst discovery for highly cost-effective and active HOR performance. However, single-atom catalysts (SACs) for the alkaline HOR have rarely been reported, and fundamental studies on the rational design of SACs are still required. Herein, the design of Ru SAC supported on the tungsten carbide (Ru SA/WC1- x ) via in situ high-temperature annealing strategy is reported. The resulting Ru SA/WC1- x catalyst exhibits remarkably enhanced HOR performance in alkaline media, a level of activity that can not be achieved with carbon-supported Ru SAC. Electrochemical results and density functional theory demonstrate that promoting the hydroxyl adsorption on Ru SA/WC1- x interfaces, which is derived from the low potential of zero charge of WC1- x support, has a significant effect on enhancing the HOR performance of SACs. This enhancement leads to 5.8 and 60.1 times higher Ru mass activity of Ru SA/WC1- x than Ru nanoparticles on carbon and Ru single-atom on N-doped carbon, respectively. This work provides new insights into the design of highly active SACs for alkaline HOR.

MatGD: Materials Graph Digitizer.

We developed Material Graph Digitizer (MatGD), which is a tool for digitizing a data line from scientific graphs. The algorithm behind the tool consists of four steps: (1) identifying graphs within subfigures, (2) separating axes and data sections, (3) discerning the data lines by eliminating irrelevant graph objects and matching with the legend, and (4) data extraction and saving. From the 62,534 papers in the areas of batteries, catalysis, and metal-organic frameworks (MOFs), 501,045 figures were mined. Remarkably, our tool showcased performance with over 99% accuracy in legend marker and text detection. Moreover, its capability for data line separation stood at 66%, which is much higher compared to those of other existing figure-mining tools. We believe that this tool will be integral to collecting both past and future data from publications, and these data can be used to train various machine learning models that can enhance material predictions and new materials discovery.

Development of a Transferable Force Field between Metal-Organic Framework and Its Polymorph.

Conventionally, force fields for specific metal-organic frameworks (MOFs) are derived from quantum chemical simulations, but this method can be computationally intensive, especially in cases for large MOF structures. In this work, we devise a methodology to reduce the force field derivation costs by replacing the original MOF with a smaller polymorphic structure, with the hypothesis that the force field parameters will be transferrable among chemically identical, polymorphic MOF structures. Specifically, we demonstrate this transferability in MOF-177 structure for H2O and NH3 gas molecules and show that the force field parameters derived from a smaller polymorphic MOF-177 can be used accurately to the original MOF-177 structure. This methodology can accelerate the development of force field parameters for large porous materials, in which computational costs for conventional methods are expensive.

Enhancing Structure-Property Relationships in Porous Materials through Transfer Learning and Cross-Material Few-Shot Learning.

Porous materials have emerged as promising solutions for a wide range of energy and environmental applications. However, the asymmetric development in the field of metal-organic frameworks (MOFs) has led to a data imbalance when it comes to MOFs versus other porous materials such as covalent organic frameworks (COFs), porous polymer networks (PPNs), and zeolites. To address this issue, we introduce PMTransformer (Porous Material Transformer), a multimodal Transformer model pretrained on a vast data set of 1.9 million hypothetical porous materials, including metal-organic frameworks, covalent organic frameworks, porous polymer networks, and zeolites. PMTransformer showcases remarkable transfer learning capabilities, resulting in state-of-the-art performance in predicting various porous material properties. To address the challenge of asymmetric data aggregation, we propose cross-material few-shot learning, which leverages the synergistic effect among different porous material classes to enhance the fine-tuning performance with a limited number of examples. As a proof of concept, we demonstrate its effectiveness in predicting band gap values of COFs using the available MOF data in the training set. Moreover, we established cross-material relationships in porous materials by predicting the unseen properties of other classes of porous materials. Our approach presents a new pathway for understanding the underlying relationships among various classes of porous materials, paving the way toward a more comprehensive understanding and design of porous materials.

Structural mechanism of Escherichia coli cyanase.

Cyanase plays a vital role in the detoxification of cyanate and supplies a continuous nitrogen source for soil microbes by converting cyanate to ammonia and carbon dioxide in a bicarbonate-dependent reaction. The structures of cyanase complexed with dianion inhibitors, in conjunction with biochemical studies, suggest putative binding sites for substrates. However, the substrate-recognition and reaction mechanisms of cyanase remain unclear. Here, crystal structures of cyanase from Escherichia coli were determined in the native form and in complexes with cyanate, bicarbonate and intermediates at 1.5-1.9 Šresolution using synchrotron X-rays and an X-ray free-electron laser. Cyanate and bicarbonate interact with the highly conserved Arg96, Ser122 and Ala123 in the active site. In the presence of a mixture of cyanate and bicarbonate, three different electron densities for intermediates were observed in the cyanase structures. Moreover, the observed electron density could explain the dynamics of the substrate or product. In addition to conformational changes in the substrate-binding pocket, dynamic movement of Leu151 was observed, which functions as a gate for the passage of substrates or products. These findings provide a structural mechanism for the substrate-binding and reaction process of cyanase.

Flash-Thermal Shock Synthesis of Single Atoms in Ambient Air.

Single-atom catalysts feature interesting catalytic activity toward applications that rely on surface reactions such as electrochemical energy storage, catalysis, and gas sensors. However, conventional synthetic approaches for such catalysts require extended periods of high-temperature annealing in vacuum systems, limiting their throughput and increasing their production cost. Herein, we report an ultrafast flash-thermal shock (FTS)-induced annealing technique (temperature > 2850 °C, <10 ms duration, and ramping/cooling rates of ∼105 K/s) that operates in an ambient-air environment to prepare single-atom-stabilized N-doped graphene. Melamine is utilized as an N-doping source to provide thermodynamically favorable metal-nitrogen bonding sites, resulting in a uniform and high-density atomic distribution of single metal atoms. To demonstrate the practical utility of the single-atom-stabilized N-doped graphene produced by the FTS method, we showcased their chemiresistive gas sensing capabilities and electrocatalytic activities. Overall, the air-ambient, ultrafast, and versatile (e.g., Co, Ni, Pt, and Co-Ni dual metal) FTS method provides a general route for high-throughput, large area, and vacuum-free manufacturing of single-atom catalysts.

Three-Dimensional MoS2/MXene Heterostructure Aerogel for Chemical Gas Sensors with Superior Sensitivity and Stability.

The concept of integrating diverse functional 2D materials into a heterostructure provides platforms for exploring physics that cannot be accessed in a single 2D material. Here, physically mixing two 2D materials, MXene and MoS2, followed by freeze-drying is utilized to successfully fabricate a 3D MoS2/MXene van der Waals heterostructure aerogel. The low-temperature synthetic approach effectively suppresses significant oxidation of the Ti3C2Tx MXene and results in a hierarchical and freestanding 3D heterostructure composed of high-quality MoS2 and MXene nanosheets. Functionalization of MXene with a MoS2 catalytic layer substantially improves sensitivity and long-term stability toward detection of NO2 gas, and computational studies are coupled with experimental results to elucidate that the mechanism behind enhancements in the gas-sensing properties is effective inhibition of HNO2 formation on the MXene surface, due to the presence of MoS2. Overall, this study has a great potential for expansion of applicability to other classes of two-dimensional materials as a general synthesis method, to be applied in future fields of catalysis and electronics.

Design of New Inorganic Crystals with the Desired Composition Using Deep Learning.

New solid-state materials have been discovered using various approaches from atom substitution in density functional theory (DFT) to generative models in machine learning. Recently, generative models have shown promising performance in finding new materials. Crystal generation with deep learning has been applied in various methods to discover new crystals. However, most generative models can only be applied to materials with specific elements or generate structures with random compositions. In this work, we developed a model that can generate crystals with desired compositions based on a crystal diffusion variational autoencoder. We generated crystal structures for 14 compositions of three types of materials in different applications. The generated structures were further stabilized using DFT calculations. We found the most stable structures in the existing database for all but one composition, even though eight compositions among them were not in the data set trained in a crystal diffusion variational autoencoder. This substantiates the prospect of the generation of an extensive range of compositions. Finally, 205 unique new crystal materials with energy above hull <100 meV/atom were generated. Moreover, we compared the average formation energy of the crystals generated from five compositions, two of which were hypothetical, with that of traditional methods like atom substitution and a generative model. The generated structures had lower formation energy than those of other models, except for one composition. These results demonstrate that our approach can be applied stably in various fields to design stable inorganic materials based on machine learning.

Density Functional Theory Study of Synergistic Gas Sensing Using an Electrically Conductive Mixed Ligand Two-Dimensional Metal-Organic Framework.

Two-dimensional conductive metal-organic frameworks (2D-cMOFs) have been adopted in electrochemical sensing applications owing to their superior electrical conductivity and large surface area. Here, we performed a density functional theory (DFT) analysis to study the synergistic impact of introducing a secondary organic ligand to the 2D-cMOF system. In this study, cobalt-hexaiminobenzene (Co-HIB) and cobalt-2,3,6,7,10,11-hexaiminotriphenylene (Co-HITP) were combined to form a mixed ligand MOF named, Co-HIB-HITP. A DFT-level comparative study was designed to access stability, synergistic gas adsorption capability, and gas adsorption mechanism, important factors in sensing material development. A potential energy surface calculation predicted the structural stability of Co-HIB-HITP at larger interlayer displacements around 3.6-4.2 Å regions along the ab-plane than its unmixed states, Co-HIB and Co-HITP, indicating the tunability of the stacking mode using the mixed ligand system. Furthermore, the adsorption capabilities toward toxic gases, NH3, H2S, NO, and NO2, were investigated, and Co-HIB-HITP revealed superiority over unmixed 2D-cMOFs in H2S and NH3 gas adsorption energies by showing 158 and 170% improvement, respectively. Finally, an electron charge density analysis revealed Co-HIB-HITP's unique stacking mode and Co-metal density as contributing factors to its gas-selective synergy effect. The AB stacked layers and an intermediate metal density (5.25%) significantly improved the electrostatic interactions with H2S and NH3 by inducing a change in the chemical environment of the gas binding sites. This work proposes the dual-ligand 2D-cMOF as the promising design strategy for the next-generation sensing material.

Computational Prediction of Stacking Mode in Conductive Two-Dimensional Metal-Organic Frameworks: An Exploration of Chemical and Electrical Property Changes.

Conductive two-dimensional metal-organic frameworks (2D MOFs) have attracted interest as they induce strong charge delocalization and improve charge carrier mobility and concentration. However, characterizing their stacking mode depends on expensive and time-consuming experimental measurements. Here, we construct a potential energy surface (PES) map database for 36 2D MOFs using density functional theory (DFT) for the experimentally synthesized and non-synthesized 2D MOFs to predict their stacking mode. The DFT PES results successfully predict the experimentally synthesized stacking mode with an accuracy of 92.9% and explain the coexistence mechanism of dual stacking modes in a single compound. Furthermore, we analyze the chemical (i.e., host-guest interaction) and electrical (i.e., electronic structure) property changes affected by stacking mode. The DFT results show that the host-guest interaction can be enhanced by the transition from AA to AB stacking, taking H2S gas as a case study. The electronic band structure calculation confirms that as AB stacking displacement increases, the in-plane charge transport pathway is reduced while the out-of-plane charge transport pathway is maintained or even increased. These results indicate that there is a trade-off between chemical and electrical properties in accordance with the stacking mode.

Rapid Joule Heating Synthesis of Oxide-Socketed High-Entropy Alloy Nanoparticles as CO2 Conversion Catalysts.

The unorthodox surface chemistry of high-entropy alloy nanoparticles (HEA-NPs), with numerous interelemental synergies, helps catalyze a variety of essential chemical processes, such as the conversion of CO2 to CO, as a sustainable path to environmental remediation. However, the risk of agglomeration and phase separation in HEA-NPs during high-temperature operations are lasting issues that impede their practical viability. Herein, we present HEA-NP catalysts that are tightly sunk in an oxide overlayer for promoting the catalytic conversion of CO2 with exceptional stability and performance. We demonstrated the controlled formation of conformal oxide overlayers on carbon nanofiber surfaces via a simple sol-gel method, which facilitated a large uptake of metal precursor ions and helped to decrease the reaction temperature required for nanoparticle formation. During the rapid thermal shock synthesis process, the oxide overlayer would also impede nanoparticle growth, resulting in uniformly distributed small HEA-NPs (2.37 ± 0.78 nm). Moreover, these HEA-NPs were firmly socketed in the reducible oxide overlayer, enabling an ultrastable catalytic performance involving >50% CO2 conversion with >97% selectivity to CO for >300 h without extensive agglomeration. Altogether, we establish the rational design principles for the thermal shock synthesis of high-entropy alloy nanoparticles and offer a helpful mechanistic perspective on how the oxide overlayer impacts the nanoparticle synthesis behavior, providing a general platform for the designed synthesis of ultrastable and high-performance catalysts that could be utilized for various industrially and environmentally relevant chemical processes.

Current Status of Korean Alcohol Drinking in Accordance with the Korean Alcohol Guidelines for Moderate Drinking Based on Facial Flushing.

Since each person has a different ability to break down alcohol, it is inappropriate to apply a uniform standard to everyone when evaluating drinking status. In Korea, there has been a guideline for moderate drinking based not only on sex and age but Koreans' alcohol metabolism capabilities that can be predicted by presence of facial flushing response. So far, there have been no studies that have investigated drinking habits of Koreans in accordance with the guideline. This study tried to identify the current drinking status of Koreans according to the guideline. As a result, it was confirmed that about 1/3 of the total population was accompanied by facial flushing when drinking alcohol, and it was found that different drinking habits were shown even in the same age and gender groups according to the presence of facial flushing. It is difficult to accurately evaluate drinking habits because facial flushing has not yet been investigated in some large data or various medical examinations. In the future, it is necessary to ensure that the presence of facial flushing can be confirmed at the medical treatment or examination site so that accurate drinking habit evaluation and prevention and resolution of drinking problems can be achieved.