The effect of curcumin supplementation on glycemic indices in adults: A meta-analysis of meta-analyses.

Curcumin, an inherent polyphenolic compound, has the potential to influence glycemic indices. Nevertheless, the conclusions drawn from extant meta-analyses remain contentious. To determine the impact of curcumin supplementation on these indices, the current umbrella meta-analysis included existing systematic reviews and meta-analyses. A thorough systematic search was conducted using databases Embase, PubMed, WOS, Scopus, and the Cochrane Library to acquire peer-reviewed literature published before January 2024. The random-effects model was employed to conduct a meta-analysis. The present analysis incorporated a total of 22 meta-analytic studies. The findings of our study indicate that the administration of curcumin supplements leads to a significant decrease in fasting blood sugar levels (FBS) (ES: -1.63; 95 % CI: -2.36, -0.89, P<0.001; I2=88.4 %, P<0.001), homeostasis model assessment-estimated insulin resistance (HOMA-IR) (ES: -0.38; 95 % CI: -0.48, -0.28, P<0.001; I2=35.9 %, P=0.142), hemoglobin A1c (HbA1c) (ES: -0.44; 95 % CI: -0.67, -0.21, P<0.001; I2=65.0 %, P=0.014), and insulin (ES: -0.86; 95 % CI: -1.52, -0.21, P=0.010; I2=92.5 %, P<0.001). The results of this study suggest that the administration of curcumin supplements may be a beneficial intervention for enhancing glycemic indices.

Prevalence and determinants of obesity and its association with upper gastrointestinal diseases in people aged 40-69 years in Yangzhong, southeast China.

Several international epidemiological studies have established a link between obesity and upper gastrointestinal cancer (UGC), but Chinese evidence is limited. This study aimed to determine the prevalence of obesity, especially central obesity, while investigating its association with upper gastrointestinal diseases in the high-risk population of Yangzhong, a typical high-risk area for UGC in southeastern China. We conducted a cross-sectional study from November 2017 to June 2021 involving 6736 residents aged 40-69. Multivariate logistic regression was used to assess independent factors influencing overweight/obesity and central obesity. We also analyzed the relationship between obesity and upper gastrointestinal diseases using multinomial logistic regression. The prevalence of overweight, obesity, waist circumference (WC), waist-to-hip ratio (WHR), and waist-to-height ratio (WHtR)-central obesity were 40.6%, 12.0%, 49.9%, 79.4%, and 63.7%, respectively. Gender, age, smoking, tea consumption, sufficient vegetable, pickled food, spicy food, eating speed, physical activity, family history of cancer, and family history of common chronic disease were associated with overweight /obesity and central obesity. Besides, education and missing teeth were only associated with central obesity. General and central obesity were positively associated with UGC, while general obesity was negatively associated with UGC precancerous diseases. There were no significant associations between obesity and UGC precancerous lesions. Subgroup analyses showed that general and central obesity was positively associated with gastric cancer but not significantly associated with esophageal cancer. Obesity is negatively and positively associated with gastric and esophageal precancerous diseases, respectively. In conclusion, general and central obesity were at high levels in the target population in this study. Most included factors influenced overweight/obesity and central obesity simultaneously. Policymakers should urgently develop individualized measures to reduce local obesity levels according to obesity characteristics. Besides, obesity increases the risk of UGC but decreases the risk of UGC precancerous diseases, especially in the stomach. The effect of obesity on the precancerous diseases of the gastric and esophagus appears to be the opposite. No significant association between obesity and upper gastrointestinal precancerous lesions was found in the study. This finding still needs to be validated in cohort studies.

Glycyrrhizic acid and glycyrrhetinic acid loaded cyclodextrin MOFs with enhanced antibacterial and anti-inflammatory effects for accelerating diabetic wound healing.

A water stable cyclodextrin MOF (Cu-SD) was synthesized with γ-cyclodextrin derivative as organic ligand and Cu2+ as metal center to co-crystallizely load glycyrrhizic acid (GL) and glycyrrhetinic acid (GA). Cu-SD has a high drug loading capacity for GL (499.91 µg/mg) and GA (112.37 µg/mg), and the drug-loaded materials had a controlled release in different meadiums. In addition, Cu-SD and its drug loaded materials demonstrated better inhibiting α-glucosidase activity than the control drug acarbose. Furthermore, Cu-SD presented excellent antibacterial activity, and the antibacterial activity was significantly enhanced after GA and GL being encapsulated by Cu-SD. Moreover, both free and drug-loaded materials had good anti-inflammatory activities, and the anti-inflammatory effects of GL@Cu-SD and GA@Cu-SD were superior to those of their corresponding free drugs. Cu-SD, GL@Cu-SD and GA@Cu-SD demonstrated good biocompatibility and were applied to treat the wounds of diabetic rats. The experimental results showed that GL@Cu-SD and GA@Cu-SD had good promoting effects on the recovery of chronic diabetic wounds by suppressing wound inflammation.

Comparison of obesity indicators for predicting cardiovascular risk factors and multimorbidity among the Chinese population based on ROC analysis.

To date, the best obesity-related indicators (ORIs) for predicting hypertension, dyslipidaemia, Type 2 diabetes mellitus (T2DM) and multimorbidity are still controversial. This study assessed the ability of 17 ORIs [body mass index (BMI), body fat percentage (BF%), c-index, Clínica Universidad de Navarra-Body Adiposity Estimator (CUN-BAE), a body shape index (ABSI), body adiposity index (BAI), waist circumference (WC), waist-hip ratio (WHR), waist-to-height ratio (WHtR), body roundness index (BRI), abdominal volume index (AVI), triglyceride glucose index (TYG), lipid accumulation product (LAP), visceral adiposity index (VAI), Chinese visceral adiposity index (CVAI), waist triglyceride index (WTI) and cardiometabolic index (CMI)] to predict hypertension, dyslipidemia, T2DM, and multimorbidity in populations aged 40-69 years. From November 2017 to December 2022, 10,432 compliant residents participated in this study. Receiver operating characteristic curves were used to assess the ability of ORIs to predict target diseases across the whole population and genders. The DeLong test was used to analyse the heterogeneity of area under curves (AUCs). Multivariable logistic regression was used to analyse the association of ORIs with hypertension, dyslipidaemia, T2DM, and multimorbidity. The prevalence of hypertension, dyslipidaemia, T2DM, and multimorbidity was 67.46%, 39.36%, 12.54% and 63.58%, respectively. After excluding ORIs associated with the target disease components, in the whole population, CVAI (AUC = 0.656), BMI (AUC = 0.655, not significantly different from WC and AVI), CVAI (AUC = 0.645, not significantly different from LAP, CMI, WHR, and WTI), and TYG (AUC = 0.740) were the best predictor of hypertension, dyslipidemia, T2DM, and multimorbidity, respectively (all P < 0.05). In the male population, BF% (AUC = 0.677), BMI (AUC = 0.698), CMI (AUC = 0.648, not significantly different from LAP and CVAI), and TYG (AUC = 0.741) were the best predictors (all P < 0.05). In the female population, CVAI (AUC = 0.677), CUN-BAE (AUC = 0.623, not significantly different from BF%, WC, WHR, WHtR, BRI and BMI), CVAI (AUC = 0.657, not significantly different from WHR), TYG (AUC = 0.740) were the best predictors (all P < 0.05). After adjusting for all covariates, all ORIs were significantly associated with hypertension, dyslipidaemia, T2DM, and multimorbidity (all P < 0.05), except for ABSI and hypertension and BAI and T2DM, which were insignificant. Ultimately, after considering the heterogeneity of prediction of ORIs among different populations, for hypertension, BF% was the best indicator for men and CVAI for the rest of the population. The best predictors of dyslipidaemia, T2DM, and multimorbidity were BMI, CVAI and TYG, respectively. Screening for common chronic diseases in combination with these factors may help to improve the effectiveness.

γ-Cyclodextrin-metal organic framework as a carrier for trans-N-p-coumaroyltyramine: A study of drug solubability, stability, and inhibitory activity against α-glucosidase.

γ-Cyclodextrin-based metal-organic frameworks (γ-CD-MOF) were successfully synthesized using the solvent diffusion method and applied as carriers for trans-N-p-coumaroyltyramine (N-p-t-CT, NCT) to study the solubability, stability, sustained release and inhibitory activity against α-glucosidase. The solubilization effect of γ-CD-MOF on N-p-t-CT was performed using impregnation (NCT@CD-MOF-1) and co-crystallization (NCT@CD-MOF-2) methods. X-ray diffraction, scanning electron microscope (SEM), fourier transform infrared spectrometer (FTIR), and N2 adsorption/desorption were used to characterize the MOFs before and after loading NCT. The results showed that NCT@CD-MOF-2 had a better solubability for N-p-t-CT, 145.03 µg/mg of drug loading capacity could be achieved, and the solubility of NCT@CD-MOF-2 in water was 366 times higher than free N-p-t-CT. In addition, the stabilities of N-p-t-CT under temperature, UV light and pH conditions were greatly improved after encapsulation in γ-CD-MOF. Furthermore, NCT@CD-MOFs had a sustained release of N-p-t-CT over an extended period in vitro due to the primary encapsulation in pore structures. Notably, γ-CD-MOF loaded with N-p-t-CT showed superior inhibitory activity against α-glucosidase compared to free N-p-t-CT. Cytotoxicity studies demonstrated that NCT@CD-MOF-2 had low toxicity in vitro and perfect biocompatibility with HL-7702 cells, and γ-CD-MOF could reduce the toxicity of free N-p-t-CT at higher concentrations.

Machine learning-driven SERS analysis platform for rapid and accurate detection of precancerous lesions of gastric cancer.

A novel approach is proposed leveraging surface-enhanced Raman spectroscopy (SERS) combined with machine learning (ML) techniques, principal component analysis (PCA)-centroid displacement-based nearest neighbor (CDNN). This label-free approach can identify slight abnormalities between SERS spectra of gastric lesions at different stages, offering a promising avenue for detection and prevention of precancerous lesion of gastric cancer (PLGC). The agaric-shaped nanoarray substrate was prepared using gas-liquid interface self-assembly and reactive ion etching (RIE) technology to measure SERS spectra of serum from mice model with gastric lesions at different stages, and then a SERS spectral recognition model was trained and constructed using the PCA-CDNN algorithm. The results showed that the agaric-shaped nanoarray substrate has good uniformity, stability, cleanliness, and SERS enhancement effect. The trained PCA-CDNN model not only found the most important features of PLGC, but also achieved satisfactory classification results with accuracy, area under curve (AUC), sensitivity, and specificity up to 100%. This demonstrated the enormous potential of this analysis platform in the diagnosis of PLGC.

Associations between the TyG index and the ɑ-Klotho protein in middle-aged and older population relevant to diabetes mellitus in NHANES 2007-2016.

BACKGROUND: The anti-aging protein Klotho has diverse functions in antioxidative stress and energy metabolism through several pathways. While it has been reported that α-Klotho is downregulated in patients with insulin resistance (IR), the association between Klotho and IR is complex and controversial. The triglyceride-glucose (TyG) index has provided a practical method for assessing IR. With this in mind, our study aimed to investigate the relationship between the TyG index and soluble α-Klotho protein levels in US populations, both with and without diabetes mellitus. METHODS: This cross-sectional study analyzed data from middle-aged and older participants in the National Health and Nutrition Examination Survey (NHANES) conducted between 2007 and 2016. The participants were divided into two groups based on their diabetes mellitus status: those with diabetes and those without diabetes. To evaluate the relationship between the TyG index and the concentration of the α-Klotho protein in each group, a series of survey-weighted multivariable linear regression models were employed. Furthermore, to examine the association between these two variables, multivariable-adjusted restricted cubic spline curves and subgroup analysis were generated. RESULTS: The study involved 6,439 adults aged 40 years or older, with a mean age of 57.8 ± 10.9 years. Among them, 1577 (24.5%) had diabetes mellitus. A subgroup analysis indicated that the presence of diabetes significantly affected the relationship between the TyG index and the α-Klotho level. After considering all covariables, regression analysis of the participants without diabetes revealed that the α-Klotho concentration decreased by 32.35 pg/ml (95% CI: -50.07, -14.64) with each one unit increase in TyG (p < 0.001). The decline in α-Klotho levels with elevated TyG was more pronounced in the female population. In patients with diabetes mellitus, a non-linear association between the TyG index and α-Klotho was observed. There was no significant correlation observed between the two when TyG index were below 9.7. However, there was an increase in klotho levels of 106.44 pg/ml for each unit increase in TyG index above 9.7 (95% CI: 28.13, 184.74) (p = 0.008). CONCLUSION: Our findings suggested that the presence of diabetes may influence the relationship between the TyG index and soluble α-Klotho. Furthermore, there seem to be sex differences in individuals without diabetes. Further studies are necessary to validate these findings.

Silicon Hybridization for the Preparation of Room-Temperature Curing and High-Temperature-Resistant Epoxy Resin.

Specialized epoxy resin, capable of achieving room-temperature profound curing and sustaining prolonged exposure to high-temperature environments, stands as a pivotal material in modern high-end manufacturing sectors including aerospace, marine equipment fabrication, machinery production, and the electronics industry. Herein, a silicon-hybridized epoxy resin, amenable to room-temperature curing and designed for high-temperature applications, was synthesized using a sol-gel methodology with silicate esters and silane coupling agents serving as silicon sources. Resin characterization indicates a uniform distribution of silicon elements in the obtained silicon hybrid epoxy resin. In comparison to the non-hybridized epoxy resin, notable improvements are observed in room-temperature curing performance, heat resistance, and mechanical strength.

Pt(IV) Complexes in the Search for Novel Platinum Prodrugs with Promising Activity.

The kinetically inert, six coordinated, octahedral Pt(IV) complexes are termed dual-, triple-, or multi-action prodrugs based on the nature of the axially substituted ligands. These ligands are either inert or biologically active, where the nature of these axial ligands provides additional stability, synergistic biological activity or cell-targeting ability. There are many literature reports from each of these classes, mentioning the varied nature of these axial ligands. The ligands comprise drug molecules such as chlorambucil, doxorubicin, valproic acid, ethacrynic acid, biologically active chalcone, coumarin, combretastatin, non-steroidal anti-inflammatory drugs (NSAIDs) and many more, potentiating the anti-proliferative profile or reducing the side effects associated with cisplatin therapy. The targeting and non-targeting nature of these moieties exert additive or synergistic effects on the anti-cancer activity of Pt(II) moieties. Herein, we discuss the effects of these axially oriented ligands and the changes in the non-leaving am(m)ine groups and in the leaving groups on the biological activity. In this review, we have presented the latest developments in the field of Pt(IV) complexes that display promising activity with a reduced resistance profile. We have discussed the structure activity relationship (SAR) and the effects of the ligands on the biological activity of Pt(IV) complexes with cisplatin, oxaliplatin, carboplatin and the Pt core other than approved drugs. This literature work will help researchers to get an idea about Pt(IV) complexes that have been classified based on the aspects of their biological activity.

The identification of novel stroke-related sphingolipid biomarkers using UPLC-MS/MS.

BACKGROUND: Stroke is a prominent contributor to global mortality and morbidity, thus necessitating the establishment of dependable diagnostic indicators. The objective of this study was to ascertain metabolites linked to sphingolipid metabolism and assess their viability as diagnostic markers for stroke. METHODS: Two cohorts, consisting of 56 S patients and 56 healthy volunteers, were incorporated into this investigation. Metabolite data was obtained through the utilization of Ultra Performance Liquid Chromatography and Tandem Mass Spectrometry (UPLC-MS/MS). The mass spectrometry data underwent targeted analysis and quantitative evaluation utilizing the multiple reaction monitoring mode of triple quadrupole mass spectrometry. Various data analysis techniques, including Orthogonal Partial Least Squares-Discriminant Analysis (OPLS-DA), least absolute shrinkage and selection operator (LASSO) regression, Support Vector Machine (SVM), logistic regression, and Receiver Operating Characteristic (ROC) curves were employed. RESULTS: A comprehensive analysis detected a total of 129 metabolites related to sphingolipid metabolism, encompassing ceramides, 1-phosphoceramides, phytoceramides, glycosphingolipids, sphingomyelins, and sphingomyelins. The implementation of OPLS-DA analysis revealed significant disparities between individuals with stroke and controls, as it successfully identified 31 metabolites that exhibited significant differential expression between the two groups. Furthermore, functional enrichment analysis indicated the participation of these metabolites in diverse biological processes. Six metabolic markers, namely CerP(d18:1/20:3), CerP(d18:1/18:1), CerP(d18:1/18:0), CerP(d18:1/16:0), SM(d18:1/26:1), and Cer(d18:0/20:0), were successfully validated as potential diagnostic markers for stroke. The utilization of ROC analysis further confirmed their diagnostic potential, while a logistic regression model incorporating these markers demonstrated robust efficacy in distinguishing stroke patients from healthy controls. CONCLUSION: these identified metabolic markers exhibit clinical significance and hold promise as valuable tools for the diagnosis of stroke.

Rapid Formation of Intramolecular Disulfide Bridges using Light: An Efficient Method to Control the Conformation and Function of Bioactive Peptides.

Disulfide bonds are widely found in natural peptides and play a pivotal role in stabilizing their secondary structures, which are highly associated with their biological functions. Herein, we introduce a light-mediated strategy to effectively control the formation of disulfides. Our strategy is based on 2-nitroveratryl (oNv), a widely used photolabile motif, which serves both as a photocaging group and an oxidant (after photolysis). We demonstrated that irradiation of oNv-caged thiols with UV light could release free thiols that are rapidly oxidized by locally released byproduct nitrosoarene, leading to a "break-to-bond" fashion. This strategy is highlighted by the in situ restoration of the antimicrobial peptide tachyplesin I (TPI) from its external disulfide-caged analogue TPI-1. TPI-1 exhibits a distorted structure and a diminished function. However, upon irradiation, the ß-hairpin structure and membrane activity of TPI were largely restored via rapid intramolecular disulfide formation. Our study proposes a powerful method to regulate the conformation and function of peptides in a spatiotemporal manner, which has significant potential for the design of disulfide-centered light-responsive systems.

Breaking the barriers of data scarcity in drug-target affinity prediction.

Accurate prediction of drug-target affinity (DTA) is of vital importance in early-stage drug discovery, facilitating the identification of drugs that can effectively interact with specific targets and regulate their activities. While wet experiments remain the most reliable method, they are time-consuming and resource-intensive, resulting in limited data availability that poses challenges for deep learning approaches. Existing methods have primarily focused on developing techniques based on the available DTA data, without adequately addressing the data scarcity issue. To overcome this challenge, we present the Semi-Supervised Multi-task training (SSM) framework for DTA prediction, which incorporates three simple yet highly effective strategies: (1) A multi-task training approach that combines DTA prediction with masked language modeling using paired drug-target data. (2) A semi-supervised training method that leverages large-scale unpaired molecules and proteins to enhance drug and target representations. This approach differs from previous methods that only employed molecules or proteins in pre-training. (3) The integration of a lightweight cross-attention module to improve the interaction between drugs and targets, further enhancing prediction accuracy. Through extensive experiments on benchmark datasets such as BindingDB, DAVIS and KIBA, we demonstrate the superior performance of our framework. Additionally, we conduct case studies on specific drug-target binding activities, virtual screening experiments, drug feature visualizations and real-world applications, all of which showcase the significant potential of our work. In conclusion, our proposed SSM-DTA framework addresses the data limitation challenge in DTA prediction and yields promising results, paving the way for more efficient and accurate drug discovery processes.

Glucose-Triggered Release of trans-N-p-Coumaroyltyramine from Zeolitic Imidazolate Framework-8 (ZIF8) Modified with Dioscorea opposita Thunb Polysaccharide.

A Dioscorea opposita Thunb polysaccharide (DOP)-modified ZIF8 material was developed in this study, which can be used as a "smart" glucose-responsive carrier to control the slow release of drugs. The 3-aminophenylboronic acid (APBA) functionalized carboxylated long-chain polymer poly(ethylene glycol) (PEG) segments, which were first modified on the surface of ZIF8 nanoparticles with a hydrogen bond and then chemically cross-linked with DOP through a borate ester bond, leading to the drugs loaded on ZIF8 being "closed" in PBS but being "open" via taking off the DOP coating in high concentrations of glucose; thus, leakage can be prevented in the drug loaded and a glucose-triggered release can effectively result. Moreover, the materials showed good biocompatibility and the released trans-N-p-coumaroyltyramine (NCT) could work synergistically with the DOP to improve insulin resistance and promote glucose consumption in insulin-resistant HepG2 cells.

Relationship between the triglyceride-glucose index and coronary artery calcification in asymptomatic, non-diabetic patients undergoing maintenance hemodialysis.

OBJECTIVE: Coronary artery calcification (CAC) is positively and independently associated with cardiovascular disease (CVD) in patients undergoing maintenance hemodialysis (MHD). Insulin resistance is independently associated with CAC and is an important risk factor for CVD. The triglyceride-glucose (TyG) index is a reliable biomarker of insulin resistance. This cross-sectional, observational study aimed to investigate the relationship between the TyG index and CAC in asymptomatic non-diabetic patients undergoing MHD. METHODS: The quantitative coronary artery calcification score (CACS) was calculated and expressed using the Agatston score. The TyG index was calculated as ln [fasting triglyceride (mg/dL) × fasting glucose (mg/dL)/2]. Multiple Poisson regression analysis, Spearman correlation analysis, and receiver operating characteristic (ROC) curves were used to investigate the relationship between the TyG index and CAC. RESULTS: The 151 patients were divided into three groups according to the tertiles of the TyG index. With an increase in the TyG index, the CACS significantly increased (Spearman's rho = 0.414, p < 0.001). Poisson regression analysis indicated that the TyG index was independently related to the presence of CAC (prevalence ratio, 1.281 [95% confidence interval, 1.121-1.465], p < 0.001). Furthermore, ROC curve analysis showed that the TyG index was of value in predicting the CAC in asymptomatic non-diabetic patients undergoing MHD, with an area under the curve of 0.667 (p = 0.010). CONCLUSION: The TyG index is independently related to the presence of CAC in asymptomatic, non-diabetic patients undergoing MHD.

The prevalence and determinant of overweight and obesity among residents aged 40-69 years in high-risk regions for upper gastrointestinal cancer in southeast China.

Being overweight or obese is one of the public health concerns worldwide, and its prevalence is gradually increasing. Obesity has been proven to be associated with some cancers, including upper gastrointestinal cancer (UGC). However, studies on the prevalence of obesity among residents of areas at high risk of UGC in China are minimal. The aim of this study is to assess the prevalence of obesity and its influencing factors among people aged 40-69 years (high-risk population) in high-risk areas for UGC in Jiangsu Province, southeast China. This cross-sectional study involved 45,036 subjects aged 40-69 years identified in the Rural Early Diagnosis and Treatment of UGC Project database in Jiangsu Province from 2017 to 2021. Differences in prevalence across gender and age were assessed using the Chi-square test. Using a multinomial logistic regression model, we examined independent risk factors for overweight/obesity and their gender and age differences. The prevalence of overweight, obesity, and overweight/obesity varied based on the standards used: Chinese standard (42.1%, 11.9%, and 54.0%) and WHO standard (34.7%, 4.7%, and 39.4%), respectively. Being overweight was more common in men than women, while obesity was more common in women than men. Age of 50-59 years, married, household size of 7-9, drinking, soy products, pickled food, and hot food intake were positively associated with overweight/obesity. Females, 60-69 years, higher education level, household size of 4-6, annual family income of more than 60,000 CNY, smoking, and fresh fruit intake were negatively associated with overweight/obesity. Stratified analysis showed that the effects of age, education and meat, egg and dairy products on overweight/obesity were different across gender. The impact of fresh fruit and vegetables on overweight/obesity was also heterogeneous between the younger (40-59 years) and older (60-69 years) groups. In conclusion, the prevalence of overweight and obesity is high among adults aged 40-69 years from high-risk areas for UGC of Jiangsu Province, southeast China. Independent influencing factors of being overweight/obese included gender, age, marital status, education, household size, annual family income, smoking, drinking, fresh fruit, soy products, pickled food and hot food intake, and may vary by gender and age. Screening-based interventions should be considered to control obesity levels among screened participants. Besides, heterogeneity of influencing factors across subgroups could be focused on to improve intervention effectiveness.

Masked Contrastive Representation Learning for Reinforcement Learning.

In pixel-based reinforcement learning (RL), the states are raw video frames, which are mapped into hidden representation before feeding to a policy network. To improve sample efficiency of state representation learning, recently, the most prominent work is based on contrastive unsupervised representation. Witnessing that consecutive video frames in a game are highly correlated, to further improve data efficiency, we propose a new algorithm, i.e., masked contrastive representation learning for RL (M-CURL), which takes the correlation among consecutive inputs into consideration. In our architecture, besides a CNN encoder for hidden presentation of input state and a policy network for action selection, we introduce an auxiliary Transformer encoder module to leverage the correlations among video frames. During training, we randomly mask the features of several frames, and use the CNN encoder and Transformer to reconstruct them based on context frames. The CNN encoder and Transformer are jointly trained via contrastive learning where the reconstructed features should be similar to the ground-truth ones while dissimilar to others. During policy evaluation, the CNN encoder and the policy network are used to take actions, and the Transformer module is discarded. Our method achieves consistent improvements over CURL on 14 out of 16 environments from DMControl suite and 23 out of 26 environments from Atari 2600 Games. The code is available at https://github.com/teslacool/m-curl.

Effect of different mixing modes on the abundance of Microcystis in Lake Taihu under green light.

Light and hydrodynamic force are important physical factors affecting growth of Microcystis. The most recent study found that green light has good effect in inhibiting growth of Microcystis. To understand the effect of mixing modes on Microcystis under the green light, we investigated the effects of continuous mixing and intermittent mixing on the abundance of Microcystis in Taihu Lake under field conditions. The study results found that abundance of Microcystis in control, intermittent mixing group, and continuous mixing group decreased 76.62%, 40.36%, and 95.18% on day 7 compared with that on day 1 in this experiment. At the end of the experiment, abundance percentages of diatoms and green algae to total phytoplankton abundance were 1.57% and 0.48% in control, 2.32% and 0.67% in intermittent mixing group, and 22.47% and 20.27% in continuous mixing group. The results indicated that continuous mixing favored the removal of Microcystis under green light conditions and was helpful for the growth of green algae and diatoms. The results provide a new approach for the control of Microcystis blooms.

R2-DDI: relation-aware feature refinement for drug-drug interaction prediction.

Precisely predicting the drug-drug interaction (DDI) is an important application and host research topic in drug discovery, especially for avoiding the adverse effect when using drug combination treatment for patients. Nowadays, machine learning and deep learning methods have achieved great success in DDI prediction. However, we notice that most of the works ignore the importance of the relation type when building the DDI prediction models. In this work, we propose a novel R$^2$-DDI framework, which introduces a relation-aware feature refinement module for drug representation learning. The relation feature is integrated into drug representation and refined in the framework. With the refinement features, we also incorporate the consistency training method to regularize the multi-branch predictions for better generalization. Through extensive experiments and studies, we demonstrate our R$^2$-DDI approach can significantly improve the DDI prediction performance over multiple real-world datasets and settings, and our method shows better generalization ability with the help of the feature refinement design.

Satisfaction and its determinants of rural upper gastrointestinal cancer screening in China: a preliminary cross-sectional study.

OBJECTIVES: To assess participants' satisfaction with and predictors of upper gastrointestinal cancer screening in rural areas. DESIGN: Cross-sectional study. SETTING: Ten screening centres in seven cities across five provinces in China. PARTICIPANTS: Of the 452 participants screened during the survey period, 438 completed the survey (response rate: 96.90%). PRIMARY OUTCOME MEASURES: Screening satisfaction was the primary outcome and it was assessed using the Patient Satisfaction Questionnaire-18. Ordinal logistic regression analysis was used to analyse the predictors of satisfaction. RESULTS: The percentages of satisfaction with the General Satisfaction, Technical Quality, Interpersonal Manner, Communication, Financial Aspects, Time Spent With Doctor, and Convenience and Accessibility dimensions were 95.89%, 79.68%, 88.36%, 83.56%, 75.11%, 82.19% and 66.44%, respectively. Education (OR 0.25, 95% CI 0.07 to 0.90), health self-assessment (OR 15.29, 95% CI 2.86 to 81.78) and family history of cancer (OR 4.20, 95% CI 1.29 to 13.71) were associated with General Satisfaction. Residence (OR 4.31, 95% CI 1.89 to 9.81) was associated with Technical Quality. Occupation (OR 0.27, 95% CI 0.08 to 0.88), health self-assessment (OR 11.30, 95% CI 3.94 to 32.43), screening purpose (OR 0.18, 95% CI 0.03 to 0.92) and distance from the screening centre (OR 4.59, 95% CI 1.35 to 15.61) were associated with interpersonal manner. Gender (OR 1.85, 95% CI 1.02 to 3.34), residence (OR 3.23, 95% CI 1.23 to 8.53) and endoscopy in the previous year (OR 2.79, 95% CI 1.13 to 6.90) were associated with Communication. Body mass index (BMI; OR 5.06, 95% CI 1.40 to 18.25) and health self-assessment (OR 2.09, 95% CI 1.12 to 3.88) were associated with financial aspects. Gender (OR 1.90, 95% CI 1.07 to 3.38), residence (OR 3.19, 95% CI 1.30 to 7.79), BMI (OR 5.26, 95% CI 1.14 to 24.34) and health self-assessment (OR 2.14, 95% CI 1.06 to 4.34) were associated with time spent with doctor. Gender (OR 1.64, 95% CI 1.04 to 2.60) and residence (OR 3.17, 95% CI 1.46 to 6.88) were associated with convenience and accessibility. CONCLUSION: There was heterogeneity across the dimensions of satisfaction with rural upper gastrointestinal cancer screening. Project manager should prioritise improving the aspects related to the convenience and accessibility dimension. Furthermore, to improve the efficiency of potential interventions, the predictors of the various dimensions should be considered.

Discovering drug-target interaction knowledge from biomedical literature.

MOTIVATION: The interaction between drugs and targets (DTI) in human body plays a crucial role in biomedical science and applications. As millions of papers come out every year in the biomedical domain, automatically discovering DTI knowledge from biomedical literature, which are usually triplets about drugs, targets and their interaction, becomes an urgent demand in the industry. Existing methods of discovering biological knowledge are mainly extractive approaches that often require detailed annotations (e.g. all mentions of biological entities, relations between every two entity mentions, etc.). However, it is difficult and costly to obtain sufficient annotations due to the requirement of expert knowledge from biomedical domains. RESULTS: To overcome these difficulties, we explore an end-to-end solution for this task by using generative approaches. We regard the DTI triplets as a sequence and use a Transformer-based model to directly generate them without using the detailed annotations of entities and relations. Further, we propose a semi-supervised method, which leverages the aforementioned end-to-end model to filter unlabeled literature and label them. Experimental results show that our method significantly outperforms extractive baselines on DTI discovery. We also create a dataset, KD-DTI, to advance this task and release it to the community. AVAILABILITY AND IMPLEMENTATION: Our code and data are available at https://github.com/bert-nmt/BERT-DTI. SUPPLEMENTARY INFORMATION: Supplementary data are available at Bioinformatics online.