Antagonistic regulation of the gibberellic acid response during stem growth in rice.

The size of plants is largely determined by growth of the stem. Stem elongation is stimulated by gibberellic acid1-3. Here we show that internode stem elongation in rice is regulated antagonistically by an 'accelerator' and a 'decelerator' in concert with gibberellic acid. Expression of a gene we name ACCELERATOR OF INTERNODE ELONGATION 1 (ACE1), which encodes a protein of unknown function, confers cells of the intercalary meristematic region with the competence for cell division, leading to internode elongation in the presence of gibberellic acid. By contrast, upregulation of DECELERATOR OF INTERNODE ELONGATION 1 (DEC1), which encodes a zinc-finger transcription factor, suppresses internode elongation, whereas downregulation of DEC1 allows internode elongation. We also show that the mechanism of internode elongation that is mediated by ACE1 and DEC1 is conserved in the Gramineae family. Furthermore, an analysis of genetic diversity suggests that mutations in ACE1 and DEC1 have historically contributed to the selection of shorter plants in domesticated populations of rice to increase their resistance to lodging, and of taller plants in wild species of rice for adaptation to growth in deep water. Our identification of these antagonistic regulatory factors enhances our understanding of the gibberellic acid response as an additional mechanism that regulates internode elongation and environmental fitness, beyond biosynthesis and gibberellic acid signal transduction.

Wheat Ym2 originated from Aegilops sharonensis and confers resistance to soil-borne Wheat yellow mosaic virus infection to the roots.

Wheat yellow mosaic virus (WYMV) is a pathogen transmitted into its host's roots by the soil-borne vector Polymyxa graminis. Ym1 and Ym2 genes protect the host from the significant yield losses caused by the virus, but the mechanistic basis of these resistance genes remains poorly understood. Here, it has been shown that Ym1 and Ym2 act within the root either by hindering the initial movement of WYMV from the vector into the root and/or by suppressing viral multiplication. A mechanical inoculation experiment on the leaf revealed that the presence of Ym1 reduced viral infection incidence, rather than viral titer, while that of Ym2 was ineffective in the leaf. To understand the basis of the root specificity of the Ym2 product, the gene was isolated from bread wheat using a positional cloning approach. The candidate gene encodes a CC-NBS-LRR protein and it correlated allelic variation with respect to its sequence with the host's disease response. Ym2 (B37500) and its paralog (B35800) are found in the near-relatives, respectively, Aegilops sharonensis and Aegilops speltoides (a close relative of the donor of bread wheat's B genome), while both sequences, in a concatenated state, are present in several accessions of the latter species. Structural diversity in Ym2 has been generated via translocation and recombination between the two genes and enhanced by the formation of a chimeric gene resulting from an intralocus recombination event. The analysis has revealed how the Ym2 region has evolved during the polyploidization events leading to the creation of cultivated wheat.

Regulator of Awn Elongation 3, an E3 ubiquitin ligase, is responsible for loss of awns during African rice domestication.

Two species of rice have been independently domesticated from different ancestral wild species in Asia and Africa. Comparison of mutations that underlie phenotypic and physiological alterations associated with domestication traits in these species gives insights into the domestication history of rice in both regions. Asian cultivated rice, Oryza sativa, and African cultivated rice, Oryza glaberrima, have been modified and improved for common traits beneficial for humans, including erect plant architecture, nonshattering seeds, nonpigmented pericarp, and lack of awns. Independent mutations in orthologous genes associated with these traits have been documented in the two cultivated species. Contrary to this prevailing model, selection for awnlessness targeted different genes in O. sativa and O. glaberrima. We identify Regulator of Awn Elongation 3 (RAE3) a gene that encodes an E3 ubiquitin ligase and is responsible for the awnless phenotype only in O. glaberrima. A 48-bp deletion may disrupt the substrate recognition domain in RAE3 and diminish awn elongation. Sequencing analysis demonstrated low nucleotide diversity in a ~600-kb region around the derived rae3 allele on chromosome 6 in O. glaberrima compared with its wild progenitor. Identification of RAE3 sheds light on the molecular mechanism underlying awn development and provides an example of how selection on different genes can confer the same domestication phenotype in Asian and African rice.

Rolapitant treats lung cancer by targeting deubiquitinase OTUD3.

BACKGROUND: Lung cancer is cancer with the highest morbidity and mortality in the world and poses a serious threat to human health. Therefore, discovering new treatments is urgently needed to improve lung cancer prognosis. Small molecule inhibitors targeting the ubiquitin-proteasome system have achieved great success, in which deubiquitinase inhibitors have broad clinical applications. The deubiquitylase OTUD3 was reported to promote lung tumorigenesis by stabilizing oncoprotein GRP78, implying that inhibition of OTUD3 may be a therapeutic strategy for lung cancer. RESULTS: In this study, we identified a small molecule inhibitor of OTUD3, Rolapitant, by computer-aided virtual screening and biological experimental verification from FDA-approved drugs library. Rolapitant inhibited the proliferation of lung cancer cells by inhibiting deubiquitinating activity of OTUD3. Quantitative proteomic profiling indicated that Rolapitant significantly upregulated the expression of death receptor 5 (DR5). Rolapitant also promoted lung cancer cell apoptosis through upregulating cell surface expression of DR5 and enhanced TRAIL-induced apoptosis. Mechanistically, Rolapitant directly targeted the OTUD3-GRP78 axis to trigger endoplasmic reticulum (ER) stress-C/EBP homologous protein (CHOP)-DR5 signaling, sensitizing lung cancer cells to TRAIL-induced apoptosis. In the vivo assays, Rolapitant suppressed the growth of lung cancer xenografts in immunocompromised mice at suitable dosages without apparent toxicity. CONCLUSION: In summary, the present study identifies Rolapitant as a novel inhibitor of deubiquitinase OTUD3 and establishes that the OTUD3-GRP78 axis is a potential therapeutic target for lung cancer.

Unveiling the Role of Electrostatic Forces on Attraction between Opposing Polyelectrolyte Brushes.

Electrostatic interaction and molecular excluded-volume effects are responsible for a plethora of nonintuitive phenomena in soft-matter systems, including local charge inversion and attraction between similar charges. In the current work, we study the surface forces and swelling behavior of opposing polyelectrolyte brushes using a classical density functional theory that accounts for electrostatic and excluded-volume correlations. We observe that the detachment pressure between similarly charged brushes is sensitive to salt concentration in both the osmotic and salted regimes and can be negative in the presence of multivalent counterions. A comparison of the theoretical results with the mean-field predictions unravels the role of correlation effects in determining the surface forces and brush structure. For systems containing multivalent counterions, the detachment pressure attains negative values at an intermediate brush-brush separation, and the attractive region in the pressure vs distance plot is magnified in terms of both the depth and width of attraction with increasing counterion valency. However, the interbrush attraction vanishes when the size-induced correlations are switched off. We also investigated the role of counterion size and polymer chain length on the detachment pressure. It is found that smaller counterions are more effective in neutralizing the polymer charge than bigger counterions, leading to a reduced interbrush repulsion and, in some cases, attraction between like-charged brushes at intermediate distances. Meanwhile, varying the chain length of the grafted polymers only shifts the location of the attraction basin, with little influence on the interaction strength. The theoretical predictions show qualitative agreement with experimental observations and offer valuable insights into the interaction between similarly charged polymer brushes in the presence of multivalent ions.

Understanding the Electric Double-Layer Structure, Capacitance, and Charging Dynamics.

Significant progress has been made in recent years in theoretical modeling of the electric double layer (EDL), a key concept in electrochemistry important for energy storage, electrocatalysis, and multitudes of other technological applications. However, major challenges remain in understanding the microscopic details of the electrochemical interface and charging mechanisms under realistic conditions. This review delves into theoretical methods to describe the equilibrium and dynamic responses of the EDL structure and capacitance for electrochemical systems commonly deployed for capacitive energy storage. Special emphasis is given to recent advances that intend to capture the nonclassical EDL behavior such as oscillatory ion distributions, polarization of nonmetallic electrodes, charge transfer, and various forms of phase transitions in the micropores of electrodes interfacing with an organic electrolyte or ionic liquid. This comprehensive analysis highlights theoretical insights into predictable relationships between materials characteristics and electrochemical performance and offers a perspective on opportunities for further development toward rational design and optimization of electrochemical systems.

CLASRP oncogene as a novel target for colorectal cancer.

Clk4-associated serine/arginine-rich protein (CLASRP), an alternative splicing regulator, may be involved in the development and progression of cancer by regulating the activity of the CDC-like kinase (Clk) family. This study explored the biological function of CLASRP in colorectal cancer (CRC). The expression of CLASRP, which is associated with clinicopathological features, was analysed in CRC tissues and paired noncancer tissues by RT-PCR. The roles of CLASRP were investigated in CRC cells transfected with plasmids or shRNA through proliferation, migration and invasion assays in vitro and a xenograft model in vivo. Apoptosis was analysed using CLASRP-overexpressing CRC cells by western blotting. Clk inhibitors were used to perform functional research on CLASRP in CLASRP-overexpressing CRC cells. CLASRP was significantly upregulated in CRC cell lines, while high CLASRP expression was correlated with metastasis in CRC patients. Functionally, overexpression of CLASRP significantly promoted the proliferation, migration and invasion of CRC cells in vitro and tumour growth in vivo. Mechanistically, the proliferation, migration and invasion of CLASRP-overexpressing CRC cells were inhibited by Clk inhibitors, accompanied by low expression of CLASRP at the gene and protein levels. Clk inhibitors induced apoptosis of CLASRP-overexpressing CRC cells, resulting in direct blockade of cell growth. The expression levels of cleaved caspase 3 and cleaved caspase 8 were increased in CLASRP-overexpressing CRC cells treated with Clk inhibitors. CLASRP might serve as a promotional oncogene in CRC cells and be suppressed by Clk inhibitors through activation of caspase pathways.

The FACT-targeted drug CBL0137 enhances the effects of rituximab to inhibit B-cell non-Hodgkin's lymphoma tumor growth by promoting apoptosis and autophagy.

BACKGROUND: Aggressive B-cell non-Hodgkin's lymphoma (B-NHL) patients often develop drug resistance and tumor recurrence after conventional immunochemotherapy, for which new treatments are needed. METHODS: We investigated the antitumor effects of CBL0137. In vitro, cell proliferation was assessed by CCK-8 and colony formation assay. Flow cytometry was performed to analyze cell cycle progression, apoptosis, mitochondrial depolarization, and reactive oxygen species (ROS) production. Autophagy was detected by transmission electron microscopy and mGFP-RFP-LC3 assay, while western blotting was employed to detect proteins involved in apoptosis and autophagy. RNA-sequencing was conducted to analyze the transcription perturbation after CBL0137 treatment in B-NHL cell lines. Finally, the efficacy and safety of CBL0137, rituximab, and their combination were tested in vivo. RESULTS: CBL0137, a small molecule anticancer agent that has significant antitumor effects in B-NHL. CBL0137 sequesters the FACT (facilitates chromatin transcription) complex from chromatin to produce cytotoxic effects in B-NHL cells. In addition, we discovered novel anticancer mechanisms of CBL0137. CBL0137 inhibited human B-NHL cell proliferation by inducing cell cycle arrest in S phase via the c-MYC/p53/p21 pathway. Furthermore, CBL0137 triggers ROS generation and induces apoptosis and autophagy in B-NHL cells through the ROS-mediated PI3K/Akt/mTOR and MAPK signaling pathways. Notably, a combination of CBL0137 and rituximab significantly suppressed B-NHL tumor growth in subcutaneous models, consistent with results at the cellular level in vitro. CONCLUSIONS: CBL0137 has potential as a novel approach for aggressive B-NHL, and its combination with rituximab can provide new therapeutic options for patients with aggressive B-NHL. Video Abstract.

Transferability evaluation of the deep potential model for simulating water-graphene confined system.

Machine learning potentials (MLPs) are poised to combine the accuracy of ab initio predictions with the computational efficiency of classical molecular dynamics (MD) simulation. While great progress has been made over the last two decades in developing MLPs, there is still much to be done to evaluate their model transferability and facilitate their development. In this work, we construct two deep potential (DP) models for liquid water near graphene surfaces, Model S and Model F, with the latter having more training data. A concurrent learning algorithm (DP-GEN) is adopted to explore the configurational space beyond the scope of conventional ab initio MD simulation. By examining the performance of Model S, we find that an accurate prediction of atomic force does not imply an accurate prediction of system energy. The deviation from the relative atomic force alone is insufficient to assess the accuracy of the DP models. Based on the performance of Model F, we propose that the relative magnitude of the model deviation and the corresponding root-mean-square error of the original test dataset, including energy and atomic force, can serve as an indicator for evaluating the accuracy of the model prediction for a given structure, which is particularly applicable for large systems where density functional theory calculations are infeasible. In addition to the prediction accuracy of the model described above, we also briefly discuss simulation stability and its relationship to the former. Both are important aspects in assessing the transferability of the MLP model.

Single-chain simulation of Ising density functional theory for weak polyelectrolytes.

Conventional theories of weak polyelectrolytes are either computationally prohibitive to account for the multidimensional inhomogeneity of polymer ionization in a liquid environment or oversimplistic in describing the coupling effects of ion-explicit electrostatic interactions and long-range intrachain correlations. To bridge this gap, we implement the Ising density functional theory (iDFT) for ionizable polymer systems using the single-chain-in-mean-field algorithm. The single-chain-in-iDFT (sc-iDFT) shows significant improvements over conventional mean-field methods in describing segment-level dissociation equilibrium, specific ion effects, and long-range intrachain correlations. With an explicit consideration of the fluctuations of polymer configurations and the position-dependent ionization of individual polymer segments, sc-iDFT provides a faithful description of the structure and thermodynamic properties of inhomogeneous weak polyelectrolyte systems across multiple length scales.

Root angle modifications by the DRO1 homolog improve rice yields in saline paddy fields.

The root system architecture (RSA) of crops can affect their production, particularly in abiotic stress conditions, such as with drought, waterlogging, and salinity. Salinity is a growing problem worldwide that negatively impacts on crop productivity, and it is believed that yields could be improved if RSAs that enabled plants to avoid saline conditions were identified. Here, we have demonstrated, through the cloning and characterization of qSOR1 (quantitative trait locus for SOIL SURFACE ROOTING 1), that a shallower root growth angle (RGA) could enhance rice yields in saline paddies. qSOR1 is negatively regulated by auxin, predominantly expressed in root columella cells, and involved in the gravitropic responses of roots. qSOR1 was found to be a homolog of DRO1 (DEEPER ROOTING 1), which is known to control RGA. CRISPR-Cas9 assays revealed that other DRO1 homologs were also involved in RGA. Introgression lines with combinations of gain-of-function and loss-of-function alleles in qSOR1 and DRO1 demonstrated four different RSAs (ultra-shallow, shallow, intermediate, and deep rooting), suggesting that natural alleles of the DRO1 homologs could be utilized to control RSA variations in rice. In saline paddies, near-isogenic lines carrying the qSOR1 loss-of-function allele had soil-surface roots (SOR) that enabled rice to avoid the reducing stresses of saline soils, resulting in increased yields compared to the parental cultivars without SOR. Our findings suggest that DRO1 homologs are valuable targets for RSA breeding and could lead to improved rice production in environments characterized by abiotic stress.

Antioxidative effect of Periplaneta americana extract on dextran sulfate sodium-induced ulcerative colitis through activation of the Nrf2 signal.

CONTEXT: Periplaneta americana L. (Blattariae) is used as a treatment for ulcerative colitis (UC) in Chinese traditional medicine. OBJECTIVE: To evaluate the antioxidative activity of P. americana whole body ethanol extract (PAE) on UC mice and whether glycine and proline could be used for quality control and identification of active PAE components. MATERIALS AND METHODS: NCM460 cells were pre-incubated in PAE, AA-L, AA-M, and AA-H (low, high and medium doses of proline and glycine), then treated with recombinant human TNF-α. The glutathione (GSH), malondialdehyde (MDA), superoxide dismutase (SOD) and reactive oxygen (ROS) levels were determined. UC mice were fed with water containing 2.5% dextran sulfate sodium (w/v) after pre-treatment with different doses of PAE once a day for 7 days. ELISA was used to detect the concentrations of inflammation-related factors. Colon tissues of mice were used to detect the activity of myeloperoxidase (MPO), GSH, MDA, and SOD. Histological changes were observed using H&E staining. The expression of target proteins was determined by western blotting. RESULTS: In vivo, PAE treatment reduced the DAI score more than in the model group, restoring the weight and colonic length. It also reduced the severity of colitis, and inflammatory and oxidative stress intensity. Additionally, western blotting showed that the Nrf2 pathway was activated by PAE. In vitro PAE significantly alleviated TNF-α-induced cell damage and oxidative stress, which is relevant to the activation of the Nrf2 pathway. CONCLUSIONS: PAE may relieve oxidative stress through the Nrf2 signaling pathway, and proline and glycine may be used as active components of its antioxidative stress activity.

Unraveling the Oxidation Behaviors of MXenes in Aqueous Systems by Active-Learning-Potential Molecular-Dynamics Simulation.

MXenes are 2D materials with great potential in various applications. However, the degradation of MXenes in humid environments has become a main obstacle in their practical use. Here we combine deep neural networks and an active learning scheme to develop a neural network potential (NNP) for aqueous MXene systems with ab initio precision but low cost. The oxidation behaviors of super large aqueous MXene systems are investigated systematically at nanosecond timescales for the first time. The oxidation process of MXenes is clearly displayed at the atomic level. Free protons and oxides greatly inhibit subsequent oxidation reactions, leading to the degree of oxidation of MXenes to exponentially decay with time, which is consistent with the oxidation rate of MXenes measured experimentally. Importantly, this computational study represents the first exploration of the kinetic process of oxidation of super-sized aqueous MXene systems. It opens a promising avenue for the future development of effective protection strategies aimed at controlling the stability of MXenes.

VLX1570 regulates the proliferation and apoptosis of human lung cancer cells through modulating ER stress and the AKT pathway.

The ubiquitin-proteasome system (UPS) possesses unique functions in tumorigenesis and has great potential for targeting tumours. The effectiveness of inhibitors targeting UPS in solid tumours remains to be studied. We screened a library of inhibitors targeting the ubiquitination system and found the highly potent, low-concentration inhibitor molecule VLX1570 that specifically killed lung cancer cells. VLX1570 is an inhibitor of deubiquitinating enzyme activity and has also shown potential for preclinical cancer treatment. We found that VLX1570 significantly inhibited lung cancer cells proliferation and colony formation. VLX1570 induced a G2/M cell cycle arrest by downregulating CDK1 and CyclinB1. Moreover, VLX1570 significantly promoted the mitochondrial-associated apoptosis. Mechanistically speaking, VLX1570 activated the PERK/IRE1/ATF6 pathway and induced ER stress in lung cancer cell lines. The inhibition of ER stress by tauroursodeoxycholic acid sodium or 4-phenylbutyric acid enhanced VLX1570-induced apoptosis. In addition, VLX1570 treatment led to the inactivation of Akt signalling and inhibited the proliferation of lung cancer cells by downregulating the Akt pathway. Moreover, combined treatment with VLX1570 and Afatinib or Gefitinib induced synergistic anti-lung cancer activity. Our present study demonstrated a novel therapy using VLX1570, which inhibited the proliferation and increased apoptosis in human lung cancer.

MicroRNA-381 in human cancer: Its involvement in tumour biology and clinical applications potential.

MicroRNAs (miRNAs) are small non-coding RNAs that regulate gene expression at the post-transcriptional level. MiRNAs are involved in the development and progression of a wide range of cancers. Among such cancer-associated miRNAs, miR-381 has been a major focus of research. The expression pattern and role of miR-381 vary among different cancer types. MiR-381 modulates various cellular behaviours in cancer, including proliferation, apoptosis, cell cycle progression, migration and invasion. MiR-381 is also involved in angiogenesis and lymphangiogenesis, as well as in the resistance to chemotherapy and radiotherapy. MiR-381 itself is regulated by several factors, such as long noncoding RNAs, circular RNAs and cytokines. Aberrant expression of miR-381 in blood samples indicates that it can be used as a diagnostic marker in cancer. Tissue miR-381 expression may serve as a prognostic factor for the clinicopathological characteristics of cancers and survival of patients. Metformin and icaritin regulate miR-381 expression and present anticancer properties. This review comprehensively summarizes the effect of miR-381 on tumour biological behaviours, as well as the clinical application potential of miR-381 for the treatment of cancer.

Cancer-associated fibroblasts and resistance to anticancer therapies: status, mechanisms, and countermeasures.

Cancer-associated fibroblasts (CAFs) are critical components of the tumor microenvironment (TME) with diverse functions such as extracellular matrix (ECM) remodeling, modulation of metabolism and angiogenesis, and crosstalk with both cancer cells and infiltrating immune cells by production of growth factors, cytokines, and chemokines. Within the TME milieu, CAFs exhibit morphological and functional transitions with relatively specific markers and hold tremendous potential to facilitate tumorigenesis, development, and resistance towards multiple therapeutic strategies including chemotherapy, radiotherapy, targeted therapy, anti-angiogenesis therapy, immunotherapy, and endocrine therapy. Accordingly, CAFs themselves and the downstream effectors and/or signaling pathways are potential targets for optimizing the sensitivity of anti-cancer therapies. This review aims to provide a detailed landscape of the role that CAFs play in conferring therapeutic resistance in different cancers and the underlying mechanisms. The translational and therapeutic perspectives of CAFs in the individualized treatment of malignant tumors are also discussed.

Substituent Dependence on Series of Cationic Gyroidal MOFs in utc-c Topology with High CO2 Affinity and Ultrahigh Anionic Dye Adsorption Capacity.

A substituent decorating strategy for modification of the functional cavity is of great importance in the design of metal-organic frameworks (MOFs). Herein, three new isostructural cationic MOFs, [Cu3(Xpip)2]·NO3·nH2O (Xpip stands for X-substituted phenylimidazophenanthroline, where X = adm (SCNU-2), f (SCNU-3), and none for SCNU-4), have been successfully synthesized and shown gyroidal utc-c topology and large pore sizes which can be adjusted by different substituents (-N(CH3)2, -F, and -H). Interestingly, the differences of the substituents (sizes and proton donor/acceptor) show essential effects on the adsorption abilities of carbon dioxide and dyes, where SCNU-4 exhibits the highest CO2 affinity and the biggest adsorption capacity for anionic dyes Fluorescein Sodium, and SCNU-3 adsorbs the largest amount (1503.6 mg/g) of Acid Fuchsin to date for the reported porous materials. The detailed studies in adsorption kinetics, adsorption isotherms, and theoretical calculation of the binding energies between the structures and dye molecules confirm that the electric properties of the frameworks (cationic) and substituents directed to the pore surface are two important factors dramatically affecting the selective dye adsorption.

The role of conformational dynamics in the activity of polymer-conjugated CalB in organic solvents.

Perennial interest in enzyme catalysis has been expanding its applicability from aqueous phases where enzymes are naturally evolved to organic solvents in which the majority of industrial chemical syntheses are carried out. Although conjugating an enzyme with a soluble polymer has been attempted to enhance enzyme activity in organic solvents, the underlying mechanism remains poorly understood in terms of the conformational dynamics and enzyme activity. Herein, we combine LF-NMR measurements and MD simulations to investigate the effects of polymer grafting on the conformational dynamics of CalB in organic solvents and the consequential impacts on the catalytic kinetics, using the lipase-catalyzed transesterification reaction as a model system. LF-NMR measurements confirm that conjugation with a soluble polymer improves the enzyme flexibility in organic solvents, leading to an increase in the catalytic efficiency of up to two orders of magnitude. MD simulations suggest that the conjugated enzyme samples a larger conformational space, compared to its native counterpart, validating the hypothesis that polymer motion enhances enzyme dynamics. These experimental and simulation results provide new insights for enhancing enzyme conformational dynamics and thereby catalytic kinetics in organic solvents.

Reliable emulation of complex functionals by active learning with error control.

A statistical emulator can be used as a surrogate of complex physics-based calculations to drastically reduce the computational cost. Its successful implementation hinges on an accurate representation of the nonlinear response surface with a high-dimensional input space. Conventional "space-filling" designs, including random sampling and Latin hypercube sampling, become inefficient as the dimensionality of the input variables increases, and the predictive accuracy of the emulator can degrade substantially for a test input distant from the training input set. To address this fundamental challenge, we develop a reliable emulator for predicting complex functionals by active learning with error control (ALEC). The algorithm is applicable to infinite-dimensional mapping with high-fidelity predictions and a controlled predictive error. The computational efficiency has been demonstrated by emulating the classical density functional theory (cDFT) calculations, a statistical-mechanical method widely used in modeling the equilibrium properties of complex molecular systems. We show that ALEC is much more accurate than conventional emulators based on the Gaussian processes with "space-filling" designs and alternative active learning methods. In addition, it is computationally more efficient than direct cDFT calculations. ALEC can be a reliable building block for emulating expensive functionals owing to its minimal computational cost, controllable predictive error, and fully automatic features.

Efficient force field and energy emulation through partition of permutationally equivalent atoms.

Gaussian process (GP) emulator has been used as a surrogate model for predicting force field and molecular potential, to overcome the computational bottleneck of ab initio molecular dynamics simulation. Integrating both atomic force and energy in predictions was found to be more accurate than using energy alone, yet it requires O((NM)3) computational operations for computing the likelihood function and making predictions, where N is the number of atoms and M is the number of simulated configurations in the training sample due to the inversion of a large covariance matrix. The high computational cost limits its applications to the simulation of small molecules. The computational challenge of using both gradient information and function values in GPs was recently noticed in machine learning communities, whereas conventional approximation methods may not work well. Here, we introduce a new approach, the atomized force field model, that integrates both force and energy in the emulator with many fewer computational operations. The drastic reduction in computation is achieved by utilizing the naturally sparse covariance structure that satisfies the constraints of the energy conservation and permutation symmetry of atoms. The efficient machine learning algorithm extends the limits of its applications on larger molecules under the same computational budget, with nearly no loss of predictive accuracy. Furthermore, our approach contains an uncertainty assessment of predictions of atomic forces and energies, useful for developing a sequential design over the chemical input space.