Feature Noise Boosts DNN Generalization Under Label Noise.

The presence of label noise in the training data has a profound impact on the generalization of deep neural networks (DNNs). In this study, we introduce and theoretically demonstrate a simple feature noise (FN) method, which directly adds noise to the features of training data and can enhance the generalization of DNNs under label noise. Specifically, we conduct theoretical analyses to reveal that label noise leads to weakened DNN generalization by loosening the generalization bound, and FN results in better DNN generalization by imposing an upper bound on the mutual information between the model weights and the features, which constrains the generalization bound. Furthermore, we conduct a qualitative analysis to discuss the ideal type of FN that obtains good label noise generalization. Finally, extensive experimental results on several popular datasets demonstrate that the FN method can significantly enhance the label noise generalization of state-of-the-art methods. The source codes of the FN method are available on https://github.com/zlzenglu/FN.

IFITM3 inhibits severe fever with thrombocytopenia syndrome virus entry and interacts with viral Gc protein.

Severe fever with thrombocytopenia syndrome (SFTS) is an emerging tick-borne hemorrhagic fever disease with high fatality rate of 10%-20%. Vaccines or specific therapeutic measures remain lacking. Human interferon inducible transmembrane protein 3 (hIFITM3) is a broad-spectrum antiviral factor targeting viral entry. However, the antiviral activity of hIFITM3 against SFTS virus (SFTSV) and the functional mechanism of IFITM3 remains unclear. Here we demonstrate that endogenous IFITM3 provides protection against SFTSV infection and participates in the anti-SFTSV effect of type Ⅰ and Ⅲ interferons (IFNs). IFITM3 overexpression exhibits anti-SFTSV function by blocking Gn/Gc-mediated viral entry and fusion. Further studies showed that IFITM3 binds SFTSV Gc directly and its intramembrane domain (IMD) is responsible for this interaction and restriction of SFTSV entry. Mutation of two neighboring cysteines on IMD weakens IFITM3-Gc interaction and attenuates the antiviral activity of IFITM3, suggesting that IFITM3-Gc interaction may partly mediate the inhibition of SFTSV entry. Overall, our data demonstrate for the first time that hIFITM3 plays a critical role in the IFNs-mediated anti-SFTSV response, and uncover a novel mechanism of IFITM3 restriction of SFTSV infection, highlighting the potential of clinical intervention on SFTS disease.

A comparative life cycle assessment on recycled concrete aggregates modified by accelerated carbonation treatment and traditional methods.

Recycling of construction and demolition wastes contributes to achieve carbon summit and carbon neutrality early in the construction industry. Accelerated carbonation is a promising new technology for enhancing the properties of recycled concrete aggregates (RCAs) as well as mitigating global warming. This study performed a comparative life cycle assessment on RCAs modified by accelerated carbonation treatment and traditional methods. The effect of different treatment methods on environmental impacts of concrete was evaluated. The key contributors of environmental impacts for concrete incorporating carbonated RCAs were identified. Moreover, a sensitivity analysis on the transport distance of concrete incorporating carbonated RCAs was conducted. Results demonstrated that incorporating carbonated RCAs could significantly reduce the energy demand, environmental impacts and environmental cost compared with natural aggregate concrete. Accelerated carbonation treatment exhibited greater potential than the normal two-stage crushing and heating treatment in mitigating environmental burden, especially for the global warming potential. Cement production and transportation were the primary contributors to environmental impacts of concrete incorporating carbonated RCAs. Sensitivity analysis indicated incorporating carbonated RCAs as alternatives of natural aggregates contributes to lower the environmental impacts of concrete when the natural aggregates are far from urban areas while the recycling center is near the city.

Cryo-EM structure of severe fever with thrombocytopenia syndrome virus.

The severe fever with thrombocytopenia syndrome virus (SFTSV) is a tick-borne human-infecting bunyavirus, which utilizes two envelope glycoproteins, Gn and Gc, to enter host cells. However, the structure and organization of these glycoproteins on virion surface are not yet known. Here we describe the structure of SFTSV determined by single particle reconstruction, which allows mechanistic insights into bunyavirus assembly at near-atomic resolution. The SFTSV Gn and Gc proteins exist as heterodimers and further assemble into pentameric and hexameric peplomers, shielding the Gc fusion loops by both intra- and inter-heterodimer interactions. Individual peplomers are associated mainly through the ectodomains, in which the highly conserved glycans on N914 of Gc play a crucial role. This elaborate assembly stabilizes Gc in the metastable prefusion conformation and creates some cryptic epitopes that are only accessible in the intermediate states during virus entry. These findings provide an important basis for developing vaccines and therapeutic drugs.

Characterization of the microbial community and prediction of metabolic functions in an anaerobic/oxic system with magnetic micropolystyrene as a biocarrier.

Polystyrene (PS) and magnetic polystyrene (MPS) materials have been used extensively in wastewater treatment. In this research, a 55-day anaerobic/oxic process was carried out to evaluate the effects of PS and MPS on microorganisms under aerobic and anaerobic conditions. Scanning electron microscopy results revealed differences in the entanglement state of the sludge with the biocarrier due to differences in surface morphology. High-throughput sequencing analysis showed that the microbial communities differed considerably in the presence of PS and MPS addition under both aerobic and anaerobic conditions. The highest abundance and diversity were observed in the PS reactor, with 929 observed species and a PD_whole_tree index of 91.58 under anaerobic conditions. MPS promoted the enrichment of bacteria related to nitrogen recycling such as Nitrospirota which increased from 1.13% in the seeding sludge to 3.48% and 10.07% in the aerobic reactors with PS and MPS, respectively. Moreover, advanced analysis showed that PS inhibited many microbial functions (e.g., protein export, nitrogen metabolism), and MPS alleviated this inhibition. This study provides significant insights into the microbial effects of PS and MPS and may shed light on biocarrier selection in future studies.

Multi-scale representation attention based deep multiple instance learning for gigapixel whole slide image analysis.

Recently, convolutional neural networks (CNNs) directly using whole slide images (WSIs) for tumor diagnosis and analysis have attracted considerable attention, because they only utilize the slide-level label for model training without any additional annotations. However, it is still a challenging task to directly handle gigapixel WSIs, due to the billions of pixels and intra-variations in each WSI. To overcome this problem, in this paper, we propose a novel end-to-end interpretable deep MIL framework for WSI analysis, by using a two-branch deep neural network and a multi-scale representation attention mechanism to directly extract features from all patches of each WSI. Specifically, we first divide each WSI into bag-, patch- and cell-level images, and then assign the slide-level label to its corresponding bag-level images, so that WSI classification becomes a MIL problem. Additionally, we design a novel multi-scale representation attention mechanism, and embed it into a two-branch deep network to simultaneously mine the bag with a correct label, the significant patches and their cell-level information. Extensive experiments demonstrate the superior performance of the proposed framework over recent state-of-the-art methods, in term of classification accuracy and model interpretability. All source codes are released at: https://github.com/xhangchen/MRAN/.

Comprehensive Evaluation of the Performance and Benefits of SSA-GGBS Geopolymer Mortar.

The activity of sewage sludge ash (SSA) is not high; ground granulated blast slag (GGBS) has a high calcium oxide content that can accelerate polymerization rates and exhibit better mechanical performance. In order to improve the engineering application of SSA-GGBS geopolymer, it is necessary to conduct a comprehensive evaluation of its performance and benefits. In this study, the fresh properties, mechanical performance and benefits of geopolymer mortar with different SSA/GGBS, modulus and Na2O contents were studied. Taking the economic and environmental benefits, working performance and mechanical performance of mortar as evaluation indexes, the entropy weight TOPSIS (Technique for Order Performance by Similarity to Ideal Solution) comprehensive evaluation method is used to evaluate the geopolymer mortar with different proportions. The results show that as SSA/GGBS increases, the workability of mortar decreases, the setting time first increases and then decreases, and the compressive strength and flexural strength decrease. By appropriately increasing the modulus, the workability of the mortar decreases and more silicates are introduced, resulting in increased strength in the later stage. By appropriately increasing the Na2O content, the volcanic ash activity of SSA and GGBS is better stimulated, the polymerization reaction is accelerated, and the early strength increases. The highest Ic (integrated cost index, Ctfc28) of geopolymer mortar is 33.95 CNY/m3/MPa, and the lowest is 16.21 CNY/m3/MPa, which is at least 41.57% higher than that of ordinary Portland cement (OPC). The minimum Ie (embodied CO2 index, Ecfc28) is 6.24 kg/m3/MPa, rising up to 14.15 kg/m3/MPa, which is at least 21.39% lower than that of OPC. The optimal mix ratio is a water-cement ratio of 0.4, a cement-sand ratio of 1.0, SSA/GGBS of 2/8, a modulus content of 1.4, and an Na2O content of 10%.

Study on Improving Measures of Mechanical Properties of Geopolymer Materials and Its Effect on CO2 Emission.

Using construction and demolition waste composites (CDWC) and fly ash (FA) to replace cement to produce concrete can reduce CO2 emissions. However, the CDWC-based geopolymer materials have two imperfections: the compressive strength is prone to decrease with the increase of curing age (strength shrinkage) under heat curing conditions, and the strength develops slowly under ambient curing conditions. To solve the problems of these materials, on the one hand, we designed an experiment of preparing CDWC-based geopolymer concrete (CDWGC) with pretreated CDWC at different high temperatures. We analyzed the influence of different pretreatment temperatures on the mechanical properties of CDWGC through compressive strength, SEM-EDS and XRD. On the other hand, we added CaO to improve the mechanical properties of CDWC-based geopolymer paste (CDWGP) under ambient curing conditions. Further, the CO2 emissions of pretreating CDWC and adding CaO were calculated by life cycle assessment (LCA). The results show that: (1) Pretreatment of CDWC can effectively solve the problem of CDWGC strength shrinkage. (2) The compressive strength of CDWGP cured at ambient can be significantly improved by adding CaO, and the compressive strength can be increased by 180.9% when the optimum content is 3%. (3) Adding CaO had less impact on CO2 emissions, a low-carbon way to improve its strength effectively.

The degradation of high emulsified oil wastewater in an intermittently-aerated MBBR packed with magnetic polystyrene particles.

This work aims to explore the effects of the magnetic polystyrene particles (MPS) on contaminants removal of the high emulsified oil wastewater. 26 days intermittently-aerated progress illustrated that COD removal efficiency and the resistance to the shock loading was promoted in the presence of MPS. Gas chromatography (GC) results also indicated that MPS enhanced the number of organic species reduced. According to the cyclic voltammetry test, conductive MPS appeared special redox performance which was considered could to facilitate the extracellular electron transfer. Furthermore, MPS dosing accelerated the electron-transporting system (ETS) activity by 24.91% compared the control. Based on the superior performance above, the conductivity of MPS is considered to be responsible for the enhanced organic removal efficiency. Moreover, the high-throughput sequencing displayed that electroactive Cloacibacterium and Acinetobacter accounted for a higher proportion in the MPS reactor. Additionally, Porphyrobacter and Dysgonomonas which were capable of degrading organics were also enriched more by MPS. To sum up, MPS is a promising additive to enhance the organic substances removal for the high emulsified oil wastewater.

GRLC: Graph Representation Learning With Constraints.

Contrastive learning has been successfully applied in unsupervised representation learning. However, the generalization ability of representation learning is limited by the fact that the loss of downstream tasks (e.g., classification) is rarely taken into account while designing contrastive methods. In this article, we propose a new contrastive-based unsupervised graph representation learning (UGRL) framework by 1) maximizing the mutual information (MI) between the semantic information and the structural information of the data and 2) designing three constraints to simultaneously consider the downstream tasks and the representation learning. As a result, our proposed method outputs robust low-dimensional representations. Experimental results on 11 public datasets demonstrate that our proposed method is superior over recent state-of-the-art methods in terms of different downstream tasks. Our code is available at https://github.com/LarryUESTC/GRLC.

Adaptive Feature Projection with Distribution Alignment for Deep Incomplete Multi-view Clustering.

Incomplete multi-view clustering (IMVC) analysis, where some views of multi-view data usually have missing data, has attracted increasing attention. However, existing IMVC methods still have two issues: (1) they pay much attention to imputing or recovering the missing data, without considering the fact that the imputed values might be inaccurate due to the unknown label information, (2) the common features of multiple views are always learned from the complete data, while ignoring the feature distribution discrepancy between the complete and incomplete data. To address these issues, we propose an imputation-free deep IMVC method and consider distribution alignment in feature learning. Concretely, the proposed method learns the features for each view by autoencoders and utilizes an adaptive feature projection to avoid the imputation for missing data. All available data are projected into a common feature space, where the common cluster information is explored by maximizing mutual information and the distribution alignment is achieved by minimizing mean discrepancy. Additionally, we design a new mean discrepancy loss for incomplete multi-view learning and make it applicable in mini-batch optimization. Extensive experiments demonstrate that our method achieves the comparable or superior performance compared with state-of-the-art methods.

Enhanced bio-affinity of magnetic QD-P(St-GMA)@Fe3O4 micro-particles via surface-quaternized modification.

In this work, a kind of bio-carrier quaternized-polystyrene-polyglycidyl methacrylate@Fe3O4 (QD-P(St-GMA)@Fe3O4, QD-PSGF) micro-particles was successfully prepared by modifying PSGF micro-particles through a hydrothermal method. The quaternary ammonium group and surface structure of QD-PSGF were confirmed through several characterization methods. We directly verified the efficacy of the quaternary ammonium group in promoting microbial activity due to QD-PSGF being synthesized by a hydrothermal method without changing the surface topography and pore. The bio-affinity of QD-PSGF microspheres was evaluated by bacterial adhesion and anaerobic digestion experiments. The results showed that a little quaternary ammonium group can increase bacterial adhesion by about 2-3 times and methane production by 40%. The novel developed QD-PSGF micro-particles can be a promising material as a biofilm carrier for bio-application.

Experimental Analysis and Establishment of Strength Attenuation Model of POM Fiber Reinforced Geopolymeric Recycled Concrete under Freeze-Thaw Cycles.

Geopolymeric recycled concrete (GRC) is a new low-carbon building material that uses both construction and industrial solid waste to replace natural aggregate and cement. GRC is similar to geopolymeric concrete (GPC) in that it has good mechanical properties but needs to be improved in terms of frost resistance. Previous studies have shown that polyoxymethylene fiber (POM fiber) can improve the shrinkage and durability of concrete and is superior to other commonly used fibers. Therefore, this paper explores adding POM fiber to GRC to improve its frost resistance. In this paper, the influence of different volumes and lengths of POM fiber on the frost resistance of geopolymeric recycled concrete (PRGRC) is studied. By measuring the changes in mass loss rate, relative dynamic elastic modulus, and compressive strength of PRGRC under different cycles, the improvement effect of POM fiber on the freeze-thaw damage of GRC is analyzed, and the strength attenuation model of PRGRC is established. The results show that the increase in POM fiber content can effectively slow down the mass loss of PRGRC in the freeze-thaw cycles, the reduction rate of relative dynamic elastic modulus, and the reduction rate of compressive strength. This shows that POM fiber can effectively improve the frost resistance of PRGRC, and the effect of 6 mm POM fiber on the freeze-thaw damage of PRGRC is better than 12 mm POM fiber. According to the test results, the existing strength attenuation model is further modified, the attenuation model of PRGRC compressive strength under the freeze-thaw cycle is obtained, and the model fitting effect is good. The strengthening mechanism of POM fiber is explained by the structural relationship between POM fiber and concrete matrix in the SEM micrograph of PRGRC. The research results provide a scientific basis for the applicability of POM fiber in geopolymeric cementitious materials and improving the frost resistance of PRGRC.

Effect of calcium peroxide assisted microwave irradiation pretreatment on humus formation and microbial community in straw and dairy manure composting.

In this study, the effects of four pretreatment methods on the crystallinity of maize straw were compared, and the CaO2 assisted microwave pretreatment was selected for straw and dairy manure composting. The humification and microbial community were investigated. Results showed that the pretreatment increased the initial water-soluble carbon, which favored the microbial activity, and the CO2 release increased by 15.71%. Pretreatment promoted the lignocellulose degradation, with total degradation ratio of 37.06%. The final humic acid content was 11.39 g/kg higher than the control. Spearman correlation analysis indicated that polyphenols and amino acids were significantly related to humus formation. In addition, pretreatment rendered the Firmicutes the most dominant phylum, and increased the metabolic intensity of reducing sugar metabolism, aromatic amino acid biosynthesis and carbon fixation pathways. Redundancy analysis revealed that the dominant genus of Firmicutes was significantly positively correlated with humus, while that of Actinobacteriota was correlated with CO2 release.

IFITM3 promotes NiV envelope protein-mediated entry into MDCK cells and interacts with the fusion subunit of the F protein.

IFITM proteins are a host restriction factor with broad-spectrum antiviral activity, but the role in the paramyxovirus entry remains unclear. Nipah virus (NiV) is a zoonotic virus of the paramyxoviridae with extremely high lethality. Here, we assessed the role of IFITM3 on NiV G and F glycoprotein-mediated virus entry. Using NiV pseudovirus bearing NiV G and F proteins to infect IFITM3-induced MDCK cells, we found that overexpression of IFITM3 promotes NiV G and F proteins-mediated virus entry. Mechanistically, the subcellular distribution showed that F protein completely co-localized with IFITM3, but G protein does not. Immunoprecipitation further indicated that IFITM3 strongly captures F protein rather than G protein. F protein truncation found that the F1 subunit completely co-localized and captures with IFITM3, but not the F2 subunit. Furthermore, IFITM3 strongly binds to F1 truncations containing fusion peptide (FP), and F1 strongly captures IFITM3 truncation with the intramembrane domain (IMD). Together, the results suggest that IFITM3 can promote NiV G and F proteins-mediated virus entry into MDCK cells, and IFITM3 directly interacts with the F1 subunit of NiV F protein dependent on the former's IMD and the latter's FP, which may occur after incorporation of fusion peptides into the cell membrane following virus fusion activation.

Traditional Chinese medicine based on Tongjiang methodology combined with proton pump inhibitor (PPI) step-down in treating non-erosive reflux disease: a study protocol for a multicentered, randomized controlled clinical trial.

BACKGROUND: Non-erosive reflux disease (NERD) is characterized by typical gastroesophageal reflux symptoms, such as heartburn and regurgitation but an absence of esophageal mucosal damage during upper gastrointestinal endoscopy. Although proton pump inhibitors (PPIs) are the first line therapy, almost 50% of patients with NERD fail to respond to this treatment. Traditional Chinese medicine (TCM) can better relieve the symptoms of NERD. Therefore, a randomized controlled trial (RCT) was designed to investigate the efficiency of TCM granules based on Tongjiang (TJ) methodology combined with PPI step-down therapy for NERD patients who did not respond to PPIs alone. METHOD: This multicentered, double-blinded, RCT with two parallel groups will recruit 174 participants who will be randomized into the TCM granules combined with PPI step-down group (n = 87) and the TCM granules placebo combined with PPI step-down group (n = 87). Both groups of participants will receive 6 weeks of treatment and 4 weeks of follow-up, and all participants will be assessed for related symptoms, mental health status, and quality of life at each visit. The primary outcome measurements include visual analog scale (VAS) for heartburn and regurgitation and the major symptoms scale. The secondary outcome measurements include PPI withdrawal rate, symptom recurrence rate, minor symptoms scale, SF-36, PRO, SAS, SDS, GERD-HRQL, and TCM syndromes scales. DISCUSSION: Previous research has shown that TCM is capable to alleviate NERD symptoms. This trial will help to provide a better understanding of the synergistic efficiency of the combination of TCM and PPIs, to explore whether the dosage of PPIs can be reduced after the supplement of TCM granules and to provide a feasible plan to reduce dependencies or withdraw NERD patients from PPIs. The outcome of this trial is expected to reduce the symptom recurrence rates, lessen patients' physical and psychological burdens, and achieve good social benefits. TRIAL REGISTRATION: Clinicaltrials.gov NCT04340297. Registered on April 9, 2020.

Improved fertilization efficiency of manure compost obtained by using oil shale semicoke as the bulking agent.

Oil shale semicoke is a kind of solid waste produced during the retorting process of oil shale, which could cause environmental pollution without reasonable disposing. In our previous study, the abandoned semicoke was recycled as bulking agent to reduce the nitrogen loss and greenhouse gases emission during composting. But influences of the obtained semicoke-blended compost on soil properties and plant growth remained unclear, which would be discussed in this study. Through leaching experiments, it was found that the N/P/K retention capacity of soil mixed with semicoke-blended compost significantly increased for the good nutrients sorption capacity of oil shale semicoke. Subsequently, germination test showed the germination index of semicoke-blended compost could attain 120%, implying its low phytotoxicity. And pot experiments exhibited the biomass of cress and Brassica rapa significantly increased by 2-4 times when applying semicoke-blended compost as fertilizer, exhibiting its great benefits to plants. For the increase of crop yield, it was closely related to their elevated nutrients uptake efficiency, also might be related to the improved soil microbial community and activity as the microbial analysis indicated. Finally, results of pollutant detection showed the concentration of polycyclic aromatic hydrocarbons, Cr, As, Cd and Pb in the mature semicoke-blended compost obtained through composting was 2.82, 95.30, 5.95, 0.34 and 14.45 mg kg-1 respectively, meeting the standard for soil application. The research suggests composting could be an effective method for the harmless disposing and resource recycling of oil shale semicoke waste.

Enhancement of biological nitrogen removal performance from low C/N municipal wastewater using novel carriers based on the nano-Fe3O4.

The aim of this work was to study the effects of the magnetic microparticles (MMP) on nitrogen removal under low C/N conditions. A 30-day anaerobic/oxic progress illustrated that nitrification and denitrification were promoted in the presence of MMP. MMP could facilitate the production of extracellular polymeric substances (EPS) and act as pH buffering in aerobic conditions. The high-throughput sequencing displayed that, compared with the sludge without MMP, the relative abundance of Dokdonella and Comamonas which are capable of both nitrifying and denitrifying were 8.7% and 1.29% higher in anaerobic sludge and 7.11% and 0.97% higher in aerobic sludge with MMP, respectively. The relative abundance of Pseudomonas with the excellent capability of EPS secretion was also observed 4.33 times higher than that without MMP in the aerobic sludge. Based on the superior performance above, MMP is a promising additive to enhance nitrogen removal efficiency for low C/N wastewater.

Factors Associated with the Magnitude Of acUpuncture treatment effectS (FAMOUS): a meta-epidemiological study of acupuncture randomised controlled trials.

OBJECTIVE: To identify factors and assess to what extent they impact the magnitude of the treatment effect of acupuncture therapies across therapeutic areas. DATA SOURCE: Medline, Embase, Cochrane Central Register of Controlled Trials, China National Knowledge Infrastructure, Wanfang Database, VIP Database, and China Biology Medicine disc, between 2015 and 2019. STUDY SELECTION: The inclusion criteria were trials with a total number of randomised patients larger than 100, at least one patient-important outcome and one of two sets of comparisons. DATA ANALYSIS: The potential independent variables were identified by reviewing relevant literature and consulting with experts. We conducted meta-regression analyses with standardised mean difference (SMD) as effect estimate for the dependent variable. The analyses included univariable meta-regression and multivariable meta-regression using a three-level robust mixed model. RESULTS: 1304 effect estimates from 584 acupuncture randomised controlled trials (RCTs) were analysed. The multivariable analyses contained 15 independent variables . In the multivariable analysis, the following produced larger treatment effects of large magnitude (>0.4): quality of life (difference of adjusted SMDs 0.51, 95% CI 0.24 to 0.77), or pain (0.48, 95% CI 0.27 to 0.69), or function (0.41, 95% CI 0.21 to 0.61) vs major events. The following produced larger treatment effects of moderate magnitude (0.2-0.4): single-centred vs multicentred RCTs (0.38, 95% CI 0.10 to 0.66); penetration acupuncture vs non-penetration types of acupuncture (0.34, 95% CI 0.15 to 0.53); non-pain symptoms vs major events (0.32, 95% CI 0.12 to 0.52). The following produced larger treatment effects of small magnitude (<0.2): high vs low frequency treatment sessions (0.19, 95% CI 0.03 to 0.35); pain vs non-pain symptoms (0.16, 95% CI 0.04 to 0.27); unreported vs reported funding (0.12, 95% CI 0 to 0.25). CONCLUSION: Patients, clinicians and policy-makers should consider penetrating over non-penetrating acupuncture and more frequent treatment sessions when feasible and acceptable. When designing future acupuncture RCTs, trialists should consider factors that impact acupuncture treatment effects.