Alzheimer’s disease(AD)is the most common form of dementia,affecting over 50 million people worldwide.This figure is projected to nearly double every 20 years,reaching 82 million by 2030 and 152 million by 2050(Alzhe...Alzheimer’s disease(AD)is the most common form of dementia,affecting over 50 million people worldwide.This figure is projected to nearly double every 20 years,reaching 82 million by 2030 and 152 million by 2050(Alzheimer’s Disease International).The apolipoproteinε4(APOE4)allele is the strongest genetic risk factor for late-onset AD(after age 65 years).Apolipoprotein E,a lipid transporter,exists in three variants:ε2,ε3,andε4.APOEε2(APOE2)is protective against AD,APOEε3(APOE3)is neutral,while APOE4 significantly increases the risk.Individuals with one copy of APOE4 have a 4-fold greater risk of developing AD,and those with two copies face an 8-fold risk compared to non-carriers.Even in cognitively normal individuals,APOE4 carriers exhibit brain metabolic and vascular deficits decades before amyloid-beta(Aβ)plaques and neurofibrillary tau tangles emerge-the hallmark pathologies of AD(Reiman et al.,2001,2005;Thambisetty et al.,2010).Notably,studies have demonstrated reduced glucose uptake,or hypometabolism,in brain regions vulnerable to AD in asymptomatic middle-aged APOE4 carriers,long before clinical symptoms arise(Reiman et al.,2001,2005).展开更多
To review the existing deep learning applications for diagnosing diabetic retinopathy and retinopathy of prematurity diseases,the available public retinal databases for the diseases and apply the International Journal...To review the existing deep learning applications for diagnosing diabetic retinopathy and retinopathy of prematurity diseases,the available public retinal databases for the diseases and apply the International Journal of Medical Informatics(IJMEDI)checklist were assessed the quality of included studies;an in-depth literature search in Scopus,Web of Science,IEEE and ACM databases targeting articles from inception up to 31st January 2023 was done by two independent reviewers.In the review,26 out of 1476 articles with a total of 36 models were included.Data size and model validation were found to be challenges for most studies.Deep learning models are gaining focus in the development of medical diagnosis tools and applications.However,there seems to be a critical issue with most of the studies being published,with some not including information about data sources and data sizes which is important for their performance verification.展开更多
Various deep learning based methods have significantlyimpacted the realm of drug discovery.The development of deep learning methods for identifying novel structural types of active compounds has become an urgent chall...Various deep learning based methods have significantlyimpacted the realm of drug discovery.The development of deep learning methods for identifying novel structural types of active compounds has become an urgent challenge.In this paper,we introduce a self-supervised representation learning framework,i.e.,Geometry-based Bidirectional Encoder Representations from Transformers(GEO-BERT).GEO-BERT considers the information of atoms and chemical bonds in chemical structures as the input,and integrates the positional information of the three-dimensional conformation of the molecule for training.Specifically,GEO-BERT enhances its ability to characterize molecular structures by introducing three different positional relationships:atom-atom,bond-bond,and atom-bond.By benchmarking study,GEO-BERT has demonstrated optimal performance on multiple benchmarks.We also performed prospective study to validate the GEO-BERT model,with screening for DYRK1A inhibitors as a case.Two potent and novel DYRK1A inhibitors(IC_(50):<1μM)were ultimately discovered.Taken together,we have developed an open-source GEO-BERT model for molecular property prediction(https://github.com/drug-designer/GEO-BERT)and proved its practical utility in early-stage drug discovery.展开更多
Cardiolipins(CLs),the mitochondria-specific class of phospholipids,are crucial to energy metabolism,cristae structure,and cell apoptosis.CLs present significant challenges in lipidomics analysis due to their structura...Cardiolipins(CLs),the mitochondria-specific class of phospholipids,are crucial to energy metabolism,cristae structure,and cell apoptosis.CLs present significant challenges in lipidomics analysis due to their structural diversity with up to four fatty acyl side chains.In this study,we developed CLAN(Cardio Lipin ANalysis),a comprehensive computational pipeline designed to improve the accuracy and coverage of cardiolipin identification.CLAN integrates three innovative modules:A cardiolipin identification module that utilizes specific fragmentation rules for precise characterization of CLs and their acyl side chains;a false positives detection module that employs retention time(RT)criteria to reduce false positives;and a prediction module that constructs regression models to identify CLs lacking authentic MS/MS spectra.CLAN achieved better identification accuracy and the highest recall rate for potential CL identification compared to the existing lipid identification tools.Furthermore,we applied CLAN program to an intermittent fasting mouse model,delineating tissue-specific CL alterations across 10 tissues.Every-other-day fasting(EODF)can partially counteract the disruption of the CL atlas across multiple tissues caused by high-fat-high-sugar diet feeding,providing novel insights into mitochondrial lipid metabolism under dietary interventions.Taken together,this work not only advances CL identification methodology but also underscores CLAN's potential in comprehensive analysis of CL atlas in the EODF animal model.CLAN is freely accessible on Git Hub.展开更多
It is important for modern hospital management to strengthen medical humanistic care and build a harmonious doctor-patient relationship.Innovative applications of the big data resources of patient experience in modern...It is important for modern hospital management to strengthen medical humanistic care and build a harmonious doctor-patient relationship.Innovative applications of the big data resources of patient experience in modern hospital management facilitate hospital management to realize real-time supervision,dynamic management and s&entitle decision-making based on patients experiences.It is helping the transformation of hospital management from an administrator^perspective to a patients perspective,and from experience-driven to data-driven.The technological innovations in hospital management based on patient experience data can assist the optimization and continuous improvement of healthcare quality,therefore help to increase patient satisfaction to the medical services.展开更多
Metal–organic frameworks(MOFs)have attracted significant research interest in biomimetic catalysis.However,the modulation of the activity of MOFs by precisely tuning the coordination of metal nodes is still a signifi...Metal–organic frameworks(MOFs)have attracted significant research interest in biomimetic catalysis.However,the modulation of the activity of MOFs by precisely tuning the coordination of metal nodes is still a significant challenge.Inspired by metalloenzymes with well-defined coordination structures,a series of MOFs containing halogen-coordinated copper nodes(Cu-X MOFs,X=Cl,Br,I)are employed to elucidate their structure–activity relationship.Intriguingly,experimental and theoretical results strongly support that precisely tuning the coordination of halogen atoms directly regulates the enzyme-like activities of Cu-X MOFs by influencing the spatial configuration and electronic structure of the Cu active center.The optimal Cu–Cl MOF exhibits excellent superoxide dismutase-like activity with a specific activity one order of magnitude higher than the reported Cu-based nanozymes.More importantly,by performing enzyme-mimicking catalysis,the Cu–Cl MOF nanozyme can significantly scavenge reactive oxygen species and alleviate oxidative stress,thus effectively relieving ocular chemical burns.Mechanistically,the antioxidant and antiapoptotic properties of Cu–Cl MOF are achieved by regulating the NRF2 and JNK or P38 MAPK pathways.Our work provides a novel way to refine MOF nanozymes by directly engineering the coordination microenvironment and,more significantly,demonstrating their potential therapeutic effect in ophthalmic disease.展开更多
On the transmission line,the invasion of foreign objects such as kites,plastic bags,and balloons and the damage to electronic components are common transmission line faults.Detecting these faults is of great significa...On the transmission line,the invasion of foreign objects such as kites,plastic bags,and balloons and the damage to electronic components are common transmission line faults.Detecting these faults is of great significance for the safe operation of power systems.Therefore,a YOLOv5 target detection method based on a deep convolution neural network is proposed.In this paper,Mobilenetv2 is used to replace Cross Stage Partial(CSP)-Darknet53 as the backbone.The structure uses depth-wise separable convolution to reduce the amount of calculation and parameters;improve the detection rate.At the same time,to compensate for the detection accuracy,the Squeeze-and-Excitation Networks(SENet)attention model is fused into the algorithm framework and a new detection scale suitable for small targets is added to improve the significance of the fault target area in the image.Collect pictures of foreign matters such as kites,plastic bags,balloons,and insulator defects of transmission lines,and sort theminto a data set.The experimental results on datasets show that themean Accuracy Precision(mAP)and recall rate of the algorithm can reach 92.1%and 92.4%,respectively.At the same time,by comparison,the detection accuracy of the proposed algorithm is higher than that of other methods.展开更多
The research examines the impact of residential and non-residential demand on facility location planning by comparing results from two location models: travel-to-work (TTW) and Residential model. The TTW model conside...The research examines the impact of residential and non-residential demand on facility location planning by comparing results from two location models: travel-to-work (TTW) and Residential model. The TTW model considers short-term changes in the state of the population due to travel-to-work (non-residential demand). By contrast, the Residential model uses a static snap-shot of the population based on official census estimates (residential demand). Comparison of both models was based on a case study of Emergency Medical Services (EMS) location-allocation planning problem in Leicester and Leicestershire, England, UK. Results showed that the using a static residential demand surface to plan EMS locations overestimates actual demand coverage, compared to a non-residential demand surface. Differences in location-allocation results between the models underscore the importance of accounting for temporal changes in the state of the population when planning locations for health service facilities. The findings of the study have implications for siting of EMS, designing, and planning of EMS service catchments and allocation of prospective demand to EMS sites. The study concludes that consideration of temporal changes in the state of the population is important for reliable and efficient location-allocation planning.展开更多
Magnetic resonance imaging(MRI)plays an important role in medical diagnosis,generating petabytes of image data annually in large hospitals.This voluminous data stream requires a significant amount of network bandwidth...Magnetic resonance imaging(MRI)plays an important role in medical diagnosis,generating petabytes of image data annually in large hospitals.This voluminous data stream requires a significant amount of network bandwidth and extensive storage infrastructure.Additionally,local data processing demands substantial manpower and hardware investments.Data isolation across different healthcare institutions hinders crossinstitutional collaboration in clinics and research.In this work,we anticipate an innovative MRI system and its four generations that integrate emerging distributed cloud computing,6G bandwidth,edge computing,federated learning,and blockchain technology.This system is called Cloud-MRI,aiming at solving the problems of MRI data storage security,transmission speed,artificial intelligence(AI)algorithm maintenance,hardware upgrading,and collaborative work.The workflow commences with the transformation of k-space raw data into the standardized Imaging Society for Magnetic Resonance in Medicine Raw Data(ISMRMRD)format.Then,the data are uploaded to the cloud or edge nodes for fast image reconstruction,neural network training,and automatic analysis.Then,the outcomes are seamlessly transmitted to clinics or research institutes for diagnosis and other services.The Cloud-MRI system will save the raw imaging data,reduce the risk of data loss,facilitate inter-institutional medical collaboration,and finally improve diagnostic accuracy and work efficiency.展开更多
Ursodeoxycholic acid(UDCA)is a naturally occurring,low-toxicity,and hydrophilic bile acid(BA)in the human body that is converted by intestinal flora using primary BA.Solute carrier family 7 member 11(SLC7A11)functions...Ursodeoxycholic acid(UDCA)is a naturally occurring,low-toxicity,and hydrophilic bile acid(BA)in the human body that is converted by intestinal flora using primary BA.Solute carrier family 7 member 11(SLC7A11)functions to uptake extracellular cystine in exchange for glutamate,and is highly expressed in a variety of human cancers.Retroperitoneal liposarcoma(RLPS)refers to liposarcoma originating from the retroperitoneal area.Lipidomics analysis revealed that UDCA was one of the most significantly downregulated metabolites in sera of RLPS patients compared with healthy subjects.The augmentation of UDCA concentration(≥25 mg/mL)demonstrated a suppressive effect on the proliferation of liposarcoma cells.[15N2]-cystine and[13C5]-glutamine isotope tracing revealed that UDCA impairs cystine uptake and glutathione(GSH)synthesis.Mechanistically,UDCA binds to the cystine transporter SLC7A11 to inhibit cystine uptake and impair GSH de novo synthesis,leading to reactive oxygen species(ROS)accumulation and mitochondrial oxidative damage.Furthermore,UDCA can promote the anti-cancer effects of ferroptosis inducers(Erastin,RSL3),the murine double minute 2(MDM2)inhibitors(Nutlin 3a,RG7112),cyclin dependent kinase 4(CDK4)inhibitor(Abemaciclib),and glutaminase inhibitor(CB839).Together,UDCA functions as a cystine exchange factor that binds to SLC7A11 for antitumor activity,and SLC7A11 is not only a new transporter for BA but also a clinically applicable target for UDCA.More importantly,in combination with other antitumor chemotherapy or physiotherapy treatments,UDCA may provide effective and promising treatment strategies for RLPS or other types of tumors in a ROS-dependent manner.展开更多
Federated learning is increasingly used in energy forecasting,enabling buildings to collaboratively predict load,photovoltaic generation,and prosumption while preserving data privacy.However,this collaborative nature ...Federated learning is increasingly used in energy forecasting,enabling buildings to collaboratively predict load,photovoltaic generation,and prosumption while preserving data privacy.However,this collaborative nature introduces new vulnerabilities,as manipulations by a single participant can propagate across the network.Such attacks can undermine grid balancing,limit flexibility provision,and reduce trust in decentralized energy systems.This work presents a comprehensive study of adversarial threats and defenses in federated energy forecasting.We compare structured manipulations generated with Generative Adversarial Networks against simple random perturbations in two attack scenarios:(i)data poisoning,where corrupted training data degrade global accuracy,and(ii)backdoors,where hidden triggers distort predictions in targeted time windows.Our experiments show that poisoning can increase global forecasting errors by up to 131%,while backdoors raise local errors by up to 48%.In both cases,Generative Adversarial Network-based attacks are consistently more effective than random perturbations,with backdoors proving especially challenging to detect due to their localized effect.To mitigate these threats,we evaluate four defense strategies:weighted aggregation,larger participant clusters,local retraining,and their coordinated integration into a secure framework.Results demonstrate that these defenses substantially reduce the impact of attacks,and in some cases even improve baseline accuracy,thereby enhancing the resilience of federated energy forecasting against adversarial manipulation.展开更多
Artificial Intelligence(AI)has evolved over the past three decades from the initial pioneering stage to become a transformative force in electrocatalytic research yet is far from realizing its full potential.This revi...Artificial Intelligence(AI)has evolved over the past three decades from the initial pioneering stage to become a transformative force in electrocatalytic research yet is far from realizing its full potential.This review traces foundational applications of AI to electrocatalysis in the 1990s to highlight the integration of AI into the full catalyst development workflow in the last five years,from material design and synthesis to characterization and performance evaluation,and ultimately to knowledge extraction.Emphasis is placed on critical but often partially recognized or neglected bottlenecks:the scale gap between atomistic simulations and macroscopic performance,inverse electrocatalyst design,physical consistency and interpretability of machine learning models,automated experiments,and the scarcity of high-quality,well validated experimental data.Cutting edge solutions such as exascale computing,machine learning interatomic potentials(MLIPs),physics-informed machine learning(PIML),generative models(variational autoencoders,diffusion models,and large language models),and FAIR-compliant data are discussed.This review highlights that the progress of AI for electrocatalysis is inherently data-centric,driven by advances in data-quality,FAIR-compliant infrastructure,and data-driven workflows that connect experiment,simulations,and machine learning.Beyond technical perspectives,this review also emphasizes the importance of interdisciplinary collaboration,industrial relevance,and cautions in respect of hyping.By identifying challenges and highlighting emerging breakthroughs,this work offers a roadmap for advancing AI-driven electrocatalysis towards more predictive,interpretable,and scalable discovery.展开更多
Background:Sclerosing adenosis(SA)and breast cancer(BC)often exhibit overlapping clinical,imaging,and pathological characteristics,making them difficult to differentiate.SA may also coexist with BC(SA+BC),including du...Background:Sclerosing adenosis(SA)and breast cancer(BC)often exhibit overlapping clinical,imaging,and pathological characteristics,making them difficult to differentiate.SA may also coexist with BC(SA+BC),including ductal carcinoma in situ(SA-DCIS)and invasive breast cancer(SA-IBC),which complicates diagnosis even when core-needle biopsy(CNB)suggests SA.This study aimed to develop interpretable AI-based binary and ternary classification models that leverage clinical and imaging features to distinguish SA-only from SA+BC and to further differentiate among SA-only,SA-DCIS,and SA-IBC.Methods:We retrospectively analyzed a cohort of 726 patients with SA(January 2006 to December 2021),comprising 537 SA-only and 189 SA+BC cases(90 SA-DCIS,99 SA-IBC).Multiple machine learning algorithms-logistic regression,support vector machine,decision tree,XGBoost,and random forest-were compared using AUC,accuracy,F1-score,and C-index.Model interpretability was assessed with SHAP to elucidate feature contributions and identify key predictors.Additionally,we incorporated an independent external validation cohort consisting of 113 patients to verify the model's effectiveness.展开更多
AIM: To compare resistin mRNA expression in subcutaneous adipose tissue (SAT) and its correlation with insulin resistance (IR) in postmenopausal obese women. METHODS: A total of 68 postmenopausal women (non obese = 34...AIM: To compare resistin mRNA expression in subcutaneous adipose tissue (SAT) and its correlation with insulin resistance (IR) in postmenopausal obese women. METHODS: A total of 68 postmenopausal women (non obese = 34 and obese = 34) were enrolled for the study. The women of the two groups were age matched (49-70 years). Fasting blood samples were collected at admission and abdominal SAT was obtained during surgery for gall bladder stones or hysterectomy. Physical parameters [age, height, weight, body mass index (BMI)] were measured. Biochemical (plasma insulin and plasma glucose) parameters were estimated by enzymatic methods. RNA was isolated by the Trizol method.SAT resistin mRNA expression was done by real time- reverse transcription polymerase chain reaction (RT-PCR) by using Quanti Tect SYBR Green RT-PCR master mix. Data was analyzed using independent Student's t test, correlation and simple linear regression analysis. RESULTS: The mean weight (52.81 ± 8.04 kg vs 79.56 ± 9.91 kg; P < 0.001), BMI (20.23 ± 3.05 kg/m 2 vs 32.19 ± 4.86 kg/m 2 ; P < 0.001), insulin (8.47 ± 3.24 U/mL vs 14.67 ± 2.18 U/mL; P < 0.001), glucose (97.44 ± 11.31 mg/dL vs 109.67 ± 8.02 mg/dL; P < 0.001) and homeostasis model assessment index (2.01 ± 0.73 vs 3.96 ± 0.61; P < 0.001) were significantly higher in postmenopausal obese women compared to postmenopausal non obese women. The mean serum resistin level was also significantly higher in postmeno-pausal obese women compared to postmenopausal non obese women (9.05 ± 5.15 vs 13.92 ± 6.32, P < 0.001). Furthermore, the mean SAT resistin mRNA expression was also significantly (0.023 ± 0.008 vs 0.036 ± 0.009; P < 0.001) higher and over expressed 1.62 fold (upregulated) in postmenopausal obese women compared to postmenopausal non obese women. In postmeno-pausal obese women, the relative SAT resistin mRNA expression showed positive (direct) and significant correlation with BMI (r = 0.78, P < 0.001) and serum resistin (r = 0.76, P < 0.001). Furthermore, the SAT resistin mRNA expression in postmenopausal obese women also showed significant and direct association (r = 0.45, P < 0.01) with IR, while in postmenopausal non obese women it did not show any association (r = -0.04, P > 0.05). CONCLUSION: Increased SAT resistin mRNA expres-sion probably leads to inducing insulin resistance and thus may be associated with obesity-related disorders in postmenopausal obese women.展开更多
Inositol 1,4,5-trisphosphate receptors(IP_(3)R)-mediated calcium ion(Ca^(2+))release plays a central role in the regulation of cell survival and death.Bcl-2 limits the Ca^(2+)release function of the IP3R through a dir...Inositol 1,4,5-trisphosphate receptors(IP_(3)R)-mediated calcium ion(Ca^(2+))release plays a central role in the regulation of cell survival and death.Bcl-2 limits the Ca^(2+)release function of the IP3R through a direct or indirect mechanism.However,the two mechanisms are overwhelmingly complex and not completely understood.Here,we convert the mechanisms into a set of ordinary differential equations.We firstly simulate the time evolution of Ca^(2+)concentration under two different levels of Bcl-2 for the direct and indirect mechanism models and compare them with experimental results available in the literature.Secondly,we employ one-and two-parameter bifurcation analysis to demonstrate that Bcl-2 can suppress Ca^(2+)signal from a global point of view both in the direct and indirect mechanism models.We then use mathematical analysis to clarify that the indirect mechanism is more efficient than the direct mechanism in repressing Ca^(2+)signal.Lastly,we predict that the two mechanisms restrict Ca^(2+)signal synergistically.Together,our study provides theoretical insights into Bcl-2 regulation in IP_(3)R-mediated Ca^(2+)release,which may be instrumental for the successful development of therapies to target Bcl-2 for cancer treatment.展开更多
Hydrogel is a kind of three-dimensional crosslinked polymer material with high moisture content.However,due to the network defects of polymer gels,traditional hydrogels are usually brittle and fragile,which limits the...Hydrogel is a kind of three-dimensional crosslinked polymer material with high moisture content.However,due to the network defects of polymer gels,traditional hydrogels are usually brittle and fragile,which limits their practical applications.Herein,we present a Hofmeister effect-aided facile strategy to prepare high-performance poly(vinyl alcohol)/montmorillonite nanocomposite hydrogels.Layered montmorillonite nanosheets can not only serve as crosslinking agents to enhance the mechanical properties of the hydrogel but also promote the ion conduction.More importantly,based on the Hofmeister effect,the presence of(NH_(4))_(2)SO_(4)can endow nanocomposite hydrogels with excellent mechanical properties by affecting PVA chains'aggregation state and crystallinity.As a result,the as-prepared nanocomposite hydrogels possess unique physical properties,including robust mechanical and electrical properties.The as-prepared hydrogels can be further assembled into a high-performance flexible sensor,which can sensitively detect large-scale and small-scale human activities.The simple design concept of this work is believed to provide a new prospect for developing robust nanocomposite hydrogels and flexible devices in the future.展开更多
基金supported by National Institute on Aging(NIH-NIA)R01AG054459(to ALL).
文摘Alzheimer’s disease(AD)is the most common form of dementia,affecting over 50 million people worldwide.This figure is projected to nearly double every 20 years,reaching 82 million by 2030 and 152 million by 2050(Alzheimer’s Disease International).The apolipoproteinε4(APOE4)allele is the strongest genetic risk factor for late-onset AD(after age 65 years).Apolipoprotein E,a lipid transporter,exists in three variants:ε2,ε3,andε4.APOEε2(APOE2)is protective against AD,APOEε3(APOE3)is neutral,while APOE4 significantly increases the risk.Individuals with one copy of APOE4 have a 4-fold greater risk of developing AD,and those with two copies face an 8-fold risk compared to non-carriers.Even in cognitively normal individuals,APOE4 carriers exhibit brain metabolic and vascular deficits decades before amyloid-beta(Aβ)plaques and neurofibrillary tau tangles emerge-the hallmark pathologies of AD(Reiman et al.,2001,2005;Thambisetty et al.,2010).Notably,studies have demonstrated reduced glucose uptake,or hypometabolism,in brain regions vulnerable to AD in asymptomatic middle-aged APOE4 carriers,long before clinical symptoms arise(Reiman et al.,2001,2005).
基金Supported by DAAD,Google Research,and the Organization for Women in Science for the Developing World(OWSD).
文摘To review the existing deep learning applications for diagnosing diabetic retinopathy and retinopathy of prematurity diseases,the available public retinal databases for the diseases and apply the International Journal of Medical Informatics(IJMEDI)checklist were assessed the quality of included studies;an in-depth literature search in Scopus,Web of Science,IEEE and ACM databases targeting articles from inception up to 31st January 2023 was done by two independent reviewers.In the review,26 out of 1476 articles with a total of 36 models were included.Data size and model validation were found to be challenges for most studies.Deep learning models are gaining focus in the development of medical diagnosis tools and applications.However,there seems to be a critical issue with most of the studies being published,with some not including information about data sources and data sizes which is important for their performance verification.
基金supported by the National Natural Science Foundation of China(Grant Nos.:62173282,62472363,and 62573367)CAMS Innovation Fund for Medical Sciences(Grant No.:2021-I2M-1–069)+1 种基金the 2024 China Industrial Technology Infrastructure Public Service Platform Project(Grant No.:GN2024-31-4700)the Foreign Expert Program of State Administration of Foreign Experts Affairs(Grant No.:H20240802)。
文摘Various deep learning based methods have significantlyimpacted the realm of drug discovery.The development of deep learning methods for identifying novel structural types of active compounds has become an urgent challenge.In this paper,we introduce a self-supervised representation learning framework,i.e.,Geometry-based Bidirectional Encoder Representations from Transformers(GEO-BERT).GEO-BERT considers the information of atoms and chemical bonds in chemical structures as the input,and integrates the positional information of the three-dimensional conformation of the molecule for training.Specifically,GEO-BERT enhances its ability to characterize molecular structures by introducing three different positional relationships:atom-atom,bond-bond,and atom-bond.By benchmarking study,GEO-BERT has demonstrated optimal performance on multiple benchmarks.We also performed prospective study to validate the GEO-BERT model,with screening for DYRK1A inhibitors as a case.Two potent and novel DYRK1A inhibitors(IC_(50):<1μM)were ultimately discovered.Taken together,we have developed an open-source GEO-BERT model for molecular property prediction(https://github.com/drug-designer/GEO-BERT)and proved its practical utility in early-stage drug discovery.
基金supported by grants from the National Key Research and Development Program of China(No.2022YFE0205800)the National Natural Science Foundation of China(No.21974114)+5 种基金Major Science and Technology Special Project of Fujian Province(No.2022YZ036012)the Fundamental Research Funds for the Central Universities(No.20720220003)Project“111”sponsored by the State Bureau of Foreign Experts and Ministry of Education of China(No.BP0618017)to S.-H.LinNatural Science Foundation of Fujian Province of China(No.2022J01330)Natural Science Foundation of Xiamen City of China(No.3502Z20227208)the China Scholarship Council(No.202308350047)to J.Zeng。
文摘Cardiolipins(CLs),the mitochondria-specific class of phospholipids,are crucial to energy metabolism,cristae structure,and cell apoptosis.CLs present significant challenges in lipidomics analysis due to their structural diversity with up to four fatty acyl side chains.In this study,we developed CLAN(Cardio Lipin ANalysis),a comprehensive computational pipeline designed to improve the accuracy and coverage of cardiolipin identification.CLAN integrates three innovative modules:A cardiolipin identification module that utilizes specific fragmentation rules for precise characterization of CLs and their acyl side chains;a false positives detection module that employs retention time(RT)criteria to reduce false positives;and a prediction module that constructs regression models to identify CLs lacking authentic MS/MS spectra.CLAN achieved better identification accuracy and the highest recall rate for potential CL identification compared to the existing lipid identification tools.Furthermore,we applied CLAN program to an intermittent fasting mouse model,delineating tissue-specific CL alterations across 10 tissues.Every-other-day fasting(EODF)can partially counteract the disruption of the CL atlas across multiple tissues caused by high-fat-high-sugar diet feeding,providing novel insights into mitochondrial lipid metabolism under dietary interventions.Taken together,this work not only advances CL identification methodology but also underscores CLAN's potential in comprehensive analysis of CL atlas in the EODF animal model.CLAN is freely accessible on Git Hub.
文摘It is important for modern hospital management to strengthen medical humanistic care and build a harmonious doctor-patient relationship.Innovative applications of the big data resources of patient experience in modern hospital management facilitate hospital management to realize real-time supervision,dynamic management and s&entitle decision-making based on patients experiences.It is helping the transformation of hospital management from an administrator^perspective to a patients perspective,and from experience-driven to data-driven.The technological innovations in hospital management based on patient experience data can assist the optimization and continuous improvement of healthcare quality,therefore help to increase patient satisfaction to the medical services.
基金the National Key R&D Program of China(Grant No.2020YFA0908100)the National Nature Science Foundation(Grant Nos.12274356,82070931,and 82271045)+1 种基金Fundamental Research Funds for the Central Universities(20720220022)the 111 Project(B16029)。
文摘Metal–organic frameworks(MOFs)have attracted significant research interest in biomimetic catalysis.However,the modulation of the activity of MOFs by precisely tuning the coordination of metal nodes is still a significant challenge.Inspired by metalloenzymes with well-defined coordination structures,a series of MOFs containing halogen-coordinated copper nodes(Cu-X MOFs,X=Cl,Br,I)are employed to elucidate their structure–activity relationship.Intriguingly,experimental and theoretical results strongly support that precisely tuning the coordination of halogen atoms directly regulates the enzyme-like activities of Cu-X MOFs by influencing the spatial configuration and electronic structure of the Cu active center.The optimal Cu–Cl MOF exhibits excellent superoxide dismutase-like activity with a specific activity one order of magnitude higher than the reported Cu-based nanozymes.More importantly,by performing enzyme-mimicking catalysis,the Cu–Cl MOF nanozyme can significantly scavenge reactive oxygen species and alleviate oxidative stress,thus effectively relieving ocular chemical burns.Mechanistically,the antioxidant and antiapoptotic properties of Cu–Cl MOF are achieved by regulating the NRF2 and JNK or P38 MAPK pathways.Our work provides a novel way to refine MOF nanozymes by directly engineering the coordination microenvironment and,more significantly,demonstrating their potential therapeutic effect in ophthalmic disease.
基金Funding project:Key Project of Science and Technology Research in Colleges andUniversities of Hebei Province.Project name:MillimeterWave Radar-Based Anti-Omission Early Warning System for School Bus Personnel.Grant Number:ZD2020318,funded to author Tang XL.Sponser:Hebei Provincial Department of Education,URL:http://jyt.hebei.gov.cn/Science and Technology Research Youth Fund Project of Hebei Province Universities.Project name:Research on Defect Detection and Engineering Vehicle Tracking System for Transmission Line Scenario.Grant Number:QN2023185,funded toW.JC,member of the mentor team.Sponser:Hebei Provincial Department of Education,URL:http://jyt.hebei.gov.cn/.
文摘On the transmission line,the invasion of foreign objects such as kites,plastic bags,and balloons and the damage to electronic components are common transmission line faults.Detecting these faults is of great significance for the safe operation of power systems.Therefore,a YOLOv5 target detection method based on a deep convolution neural network is proposed.In this paper,Mobilenetv2 is used to replace Cross Stage Partial(CSP)-Darknet53 as the backbone.The structure uses depth-wise separable convolution to reduce the amount of calculation and parameters;improve the detection rate.At the same time,to compensate for the detection accuracy,the Squeeze-and-Excitation Networks(SENet)attention model is fused into the algorithm framework and a new detection scale suitable for small targets is added to improve the significance of the fault target area in the image.Collect pictures of foreign matters such as kites,plastic bags,balloons,and insulator defects of transmission lines,and sort theminto a data set.The experimental results on datasets show that themean Accuracy Precision(mAP)and recall rate of the algorithm can reach 92.1%and 92.4%,respectively.At the same time,by comparison,the detection accuracy of the proposed algorithm is higher than that of other methods.
文摘The research examines the impact of residential and non-residential demand on facility location planning by comparing results from two location models: travel-to-work (TTW) and Residential model. The TTW model considers short-term changes in the state of the population due to travel-to-work (non-residential demand). By contrast, the Residential model uses a static snap-shot of the population based on official census estimates (residential demand). Comparison of both models was based on a case study of Emergency Medical Services (EMS) location-allocation planning problem in Leicester and Leicestershire, England, UK. Results showed that the using a static residential demand surface to plan EMS locations overestimates actual demand coverage, compared to a non-residential demand surface. Differences in location-allocation results between the models underscore the importance of accounting for temporal changes in the state of the population when planning locations for health service facilities. The findings of the study have implications for siting of EMS, designing, and planning of EMS service catchments and allocation of prospective demand to EMS sites. The study concludes that consideration of temporal changes in the state of the population is important for reliable and efficient location-allocation planning.
基金supported by the National Natural Science Foundation of China(62122064,62331021,62371410)the Natural Science Foundation of Fujian Province of China(2023J02005 and 2021J011184)+1 种基金the President Fund of Xiamen University(20720220063)the Nanqiang Outstanding Talents Program of Xiamen University.
文摘Magnetic resonance imaging(MRI)plays an important role in medical diagnosis,generating petabytes of image data annually in large hospitals.This voluminous data stream requires a significant amount of network bandwidth and extensive storage infrastructure.Additionally,local data processing demands substantial manpower and hardware investments.Data isolation across different healthcare institutions hinders crossinstitutional collaboration in clinics and research.In this work,we anticipate an innovative MRI system and its four generations that integrate emerging distributed cloud computing,6G bandwidth,edge computing,federated learning,and blockchain technology.This system is called Cloud-MRI,aiming at solving the problems of MRI data storage security,transmission speed,artificial intelligence(AI)algorithm maintenance,hardware upgrading,and collaborative work.The workflow commences with the transformation of k-space raw data into the standardized Imaging Society for Magnetic Resonance in Medicine Raw Data(ISMRMRD)format.Then,the data are uploaded to the cloud or edge nodes for fast image reconstruction,neural network training,and automatic analysis.Then,the outcomes are seamlessly transmitted to clinics or research institutes for diagnosis and other services.The Cloud-MRI system will save the raw imaging data,reduce the risk of data loss,facilitate inter-institutional medical collaboration,and finally improve diagnostic accuracy and work efficiency.
基金supported by grants from the National Natural Science Foundation of China(Grant NOs.:82272935,91957120 and 21974114)the Postdoctoral Fellowship Program of CPSF(Program No.:GZC20240901)+5 种基金the Xiamen Medical Industry Combined Guidance Project,China(Project No.:3502Z20244ZD2022)the Scientific Research Foundation for Advanced Talents,Xiang'an Hospital of Xiamen University,China(Grant No.:PM20180917008)the Fundamental Research Funds for the Central Universities,China(Grant No.:20720210001)Major Science and Technology Special Project of Fujian Province,China(Project No.:2022YZ036012)Joint Laboratory of School of Medicine,Xiamen University-Shanghai Jiangxia Blood Technology Co.,Ltd.,China(Grant No.:XDHT2020010C)Joint Research Center of School of Medicine,Xiamen University-Jiangsu Charity Biotech Co.,Ltd.,China(Grant No.:20233160C0002).
文摘Ursodeoxycholic acid(UDCA)is a naturally occurring,low-toxicity,and hydrophilic bile acid(BA)in the human body that is converted by intestinal flora using primary BA.Solute carrier family 7 member 11(SLC7A11)functions to uptake extracellular cystine in exchange for glutamate,and is highly expressed in a variety of human cancers.Retroperitoneal liposarcoma(RLPS)refers to liposarcoma originating from the retroperitoneal area.Lipidomics analysis revealed that UDCA was one of the most significantly downregulated metabolites in sera of RLPS patients compared with healthy subjects.The augmentation of UDCA concentration(≥25 mg/mL)demonstrated a suppressive effect on the proliferation of liposarcoma cells.[15N2]-cystine and[13C5]-glutamine isotope tracing revealed that UDCA impairs cystine uptake and glutathione(GSH)synthesis.Mechanistically,UDCA binds to the cystine transporter SLC7A11 to inhibit cystine uptake and impair GSH de novo synthesis,leading to reactive oxygen species(ROS)accumulation and mitochondrial oxidative damage.Furthermore,UDCA can promote the anti-cancer effects of ferroptosis inducers(Erastin,RSL3),the murine double minute 2(MDM2)inhibitors(Nutlin 3a,RG7112),cyclin dependent kinase 4(CDK4)inhibitor(Abemaciclib),and glutaminase inhibitor(CB839).Together,UDCA functions as a cystine exchange factor that binds to SLC7A11 for antitumor activity,and SLC7A11 is not only a new transporter for BA but also a clinically applicable target for UDCA.More importantly,in combination with other antitumor chemotherapy or physiotherapy treatments,UDCA may provide effective and promising treatment strategies for RLPS or other types of tumors in a ROS-dependent manner.
基金KIT-Publication Fund of the Karl-sruhe Institute of Technology,GermanyGerman Research Foundation(DFG)as part of the Research Training Group 2153:Energy Status Data-Informatics Methods for its Collection,Analysis,and Exploitation.
文摘Federated learning is increasingly used in energy forecasting,enabling buildings to collaboratively predict load,photovoltaic generation,and prosumption while preserving data privacy.However,this collaborative nature introduces new vulnerabilities,as manipulations by a single participant can propagate across the network.Such attacks can undermine grid balancing,limit flexibility provision,and reduce trust in decentralized energy systems.This work presents a comprehensive study of adversarial threats and defenses in federated energy forecasting.We compare structured manipulations generated with Generative Adversarial Networks against simple random perturbations in two attack scenarios:(i)data poisoning,where corrupted training data degrade global accuracy,and(ii)backdoors,where hidden triggers distort predictions in targeted time windows.Our experiments show that poisoning can increase global forecasting errors by up to 131%,while backdoors raise local errors by up to 48%.In both cases,Generative Adversarial Network-based attacks are consistently more effective than random perturbations,with backdoors proving especially challenging to detect due to their localized effect.To mitigate these threats,we evaluate four defense strategies:weighted aggregation,larger participant clusters,local retraining,and their coordinated integration into a secure framework.Results demonstrate that these defenses substantially reduce the impact of attacks,and in some cases even improve baseline accuracy,thereby enhancing the resilience of federated energy forecasting against adversarial manipulation.
基金the support of National Natural Science Foundation of China(42307381).
文摘Artificial Intelligence(AI)has evolved over the past three decades from the initial pioneering stage to become a transformative force in electrocatalytic research yet is far from realizing its full potential.This review traces foundational applications of AI to electrocatalysis in the 1990s to highlight the integration of AI into the full catalyst development workflow in the last five years,from material design and synthesis to characterization and performance evaluation,and ultimately to knowledge extraction.Emphasis is placed on critical but often partially recognized or neglected bottlenecks:the scale gap between atomistic simulations and macroscopic performance,inverse electrocatalyst design,physical consistency and interpretability of machine learning models,automated experiments,and the scarcity of high-quality,well validated experimental data.Cutting edge solutions such as exascale computing,machine learning interatomic potentials(MLIPs),physics-informed machine learning(PIML),generative models(variational autoencoders,diffusion models,and large language models),and FAIR-compliant data are discussed.This review highlights that the progress of AI for electrocatalysis is inherently data-centric,driven by advances in data-quality,FAIR-compliant infrastructure,and data-driven workflows that connect experiment,simulations,and machine learning.Beyond technical perspectives,this review also emphasizes the importance of interdisciplinary collaboration,industrial relevance,and cautions in respect of hyping.By identifying challenges and highlighting emerging breakthroughs,this work offers a roadmap for advancing AI-driven electrocatalysis towards more predictive,interpretable,and scalable discovery.
基金National High Level Hospital Clinical Research Funding,Grant/Award Numbers:2025-PUMCH-A-147,2022-PUMCH-B-039Chinese Academy of Medical Sciences(CAMS)Innovation Fund for Medical Sciences(CIFM),Grant/Award Number:2021-I2M-1-014。
文摘Background:Sclerosing adenosis(SA)and breast cancer(BC)often exhibit overlapping clinical,imaging,and pathological characteristics,making them difficult to differentiate.SA may also coexist with BC(SA+BC),including ductal carcinoma in situ(SA-DCIS)and invasive breast cancer(SA-IBC),which complicates diagnosis even when core-needle biopsy(CNB)suggests SA.This study aimed to develop interpretable AI-based binary and ternary classification models that leverage clinical and imaging features to distinguish SA-only from SA+BC and to further differentiate among SA-only,SA-DCIS,and SA-IBC.Methods:We retrospectively analyzed a cohort of 726 patients with SA(January 2006 to December 2021),comprising 537 SA-only and 189 SA+BC cases(90 SA-DCIS,99 SA-IBC).Multiple machine learning algorithms-logistic regression,support vector machine,decision tree,XGBoost,and random forest-were compared using AUC,accuracy,F1-score,and C-index.Model interpretability was assessed with SHAP to elucidate feature contributions and identify key predictors.Additionally,we incorporated an independent external validation cohort consisting of 113 patients to verify the model's effectiveness.
基金Supported by Indian Council of Medical Research, New Delhi,IndiaCentral Council Research in Yoga and Naturopathy,New Delhi, India, to Sadashiv and Tiwari S
文摘AIM: To compare resistin mRNA expression in subcutaneous adipose tissue (SAT) and its correlation with insulin resistance (IR) in postmenopausal obese women. METHODS: A total of 68 postmenopausal women (non obese = 34 and obese = 34) were enrolled for the study. The women of the two groups were age matched (49-70 years). Fasting blood samples were collected at admission and abdominal SAT was obtained during surgery for gall bladder stones or hysterectomy. Physical parameters [age, height, weight, body mass index (BMI)] were measured. Biochemical (plasma insulin and plasma glucose) parameters were estimated by enzymatic methods. RNA was isolated by the Trizol method.SAT resistin mRNA expression was done by real time- reverse transcription polymerase chain reaction (RT-PCR) by using Quanti Tect SYBR Green RT-PCR master mix. Data was analyzed using independent Student's t test, correlation and simple linear regression analysis. RESULTS: The mean weight (52.81 ± 8.04 kg vs 79.56 ± 9.91 kg; P < 0.001), BMI (20.23 ± 3.05 kg/m 2 vs 32.19 ± 4.86 kg/m 2 ; P < 0.001), insulin (8.47 ± 3.24 U/mL vs 14.67 ± 2.18 U/mL; P < 0.001), glucose (97.44 ± 11.31 mg/dL vs 109.67 ± 8.02 mg/dL; P < 0.001) and homeostasis model assessment index (2.01 ± 0.73 vs 3.96 ± 0.61; P < 0.001) were significantly higher in postmenopausal obese women compared to postmenopausal non obese women. The mean serum resistin level was also significantly higher in postmeno-pausal obese women compared to postmenopausal non obese women (9.05 ± 5.15 vs 13.92 ± 6.32, P < 0.001). Furthermore, the mean SAT resistin mRNA expression was also significantly (0.023 ± 0.008 vs 0.036 ± 0.009; P < 0.001) higher and over expressed 1.62 fold (upregulated) in postmenopausal obese women compared to postmenopausal non obese women. In postmeno-pausal obese women, the relative SAT resistin mRNA expression showed positive (direct) and significant correlation with BMI (r = 0.78, P < 0.001) and serum resistin (r = 0.76, P < 0.001). Furthermore, the SAT resistin mRNA expression in postmenopausal obese women also showed significant and direct association (r = 0.45, P < 0.01) with IR, while in postmenopausal non obese women it did not show any association (r = -0.04, P > 0.05). CONCLUSION: Increased SAT resistin mRNA expres-sion probably leads to inducing insulin resistance and thus may be associated with obesity-related disorders in postmenopausal obese women.
基金supported by Shanxi Province Science Foundation for Youths(Grant No.201901D211159)the National Natural Science Foundation of China(Grant Nos.11504214,11874310,and 12090052).
文摘Inositol 1,4,5-trisphosphate receptors(IP_(3)R)-mediated calcium ion(Ca^(2+))release plays a central role in the regulation of cell survival and death.Bcl-2 limits the Ca^(2+)release function of the IP3R through a direct or indirect mechanism.However,the two mechanisms are overwhelmingly complex and not completely understood.Here,we convert the mechanisms into a set of ordinary differential equations.We firstly simulate the time evolution of Ca^(2+)concentration under two different levels of Bcl-2 for the direct and indirect mechanism models and compare them with experimental results available in the literature.Secondly,we employ one-and two-parameter bifurcation analysis to demonstrate that Bcl-2 can suppress Ca^(2+)signal from a global point of view both in the direct and indirect mechanism models.We then use mathematical analysis to clarify that the indirect mechanism is more efficient than the direct mechanism in repressing Ca^(2+)signal.Lastly,we predict that the two mechanisms restrict Ca^(2+)signal synergistically.Together,our study provides theoretical insights into Bcl-2 regulation in IP_(3)R-mediated Ca^(2+)release,which may be instrumental for the successful development of therapies to target Bcl-2 for cancer treatment.
基金Project supported by the National Natural Science Foundation of China(Grant No.12274356)the Fundamental Research Funds for the Central Universities(Grant No.20720220022)the 111 Project(Grant No.B16029)。
文摘Hydrogel is a kind of three-dimensional crosslinked polymer material with high moisture content.However,due to the network defects of polymer gels,traditional hydrogels are usually brittle and fragile,which limits their practical applications.Herein,we present a Hofmeister effect-aided facile strategy to prepare high-performance poly(vinyl alcohol)/montmorillonite nanocomposite hydrogels.Layered montmorillonite nanosheets can not only serve as crosslinking agents to enhance the mechanical properties of the hydrogel but also promote the ion conduction.More importantly,based on the Hofmeister effect,the presence of(NH_(4))_(2)SO_(4)can endow nanocomposite hydrogels with excellent mechanical properties by affecting PVA chains'aggregation state and crystallinity.As a result,the as-prepared nanocomposite hydrogels possess unique physical properties,including robust mechanical and electrical properties.The as-prepared hydrogels can be further assembled into a high-performance flexible sensor,which can sensitively detect large-scale and small-scale human activities.The simple design concept of this work is believed to provide a new prospect for developing robust nanocomposite hydrogels and flexible devices in the future.
