Search

As the precursor to NAD+ and NADP+, niacin is important for catabolic and anabolic redox reactions. In addition, niacin is known for its anti-lipolytic action via a hydroxycarboxylic acid-2-receptor-dependent mechanism. The anti-lipolytic effects of traditional free niacin supplementation during transition periods had been studied extensively, but the reported effects are ambiguous. In the past decade, a series of studies were conducted to evaluate the effects of rumen-protected niacin (RPN) on production performance and metabolic status in early lactation and on heat stress in dairy cows. Feeding RPN seems more effective than free niacin regarding increasing circulating niacin concentration. The rebound of plasma NEFA was found after termination of niacin abomasal infusion. Feeding RPN or infusion of niacin via the abomasum could suppress lipolysis and reduce insulin resistance in early lactation. Additionally, RPN supplementation could possibly relieve heat stress through vasodilation during moderate to severe heat stress condition. However, these beneficial effects of niacin supplementation have not always been observed. The inconsistent results across studies may be related to dosages of niacin supplementation, rebound of plasma NEFA concentration, stage of lactation or severity of heat stress. Overall, the current review is to present updated information on niacin nutrition in dairy cows and the recommendations are given for future research.

In this paper, a novel Deep Semantic Space learning model with Intra-class Low-rank constraint (DSSIL) is proposed for cross-modal retrieval, which is composed of two subnetworks for modality-specific representation learning, followed by projection layers for common space mapping. In particular, DSSIL takes into account semantic consistency to fuse the cross-modal data in a high-level common space, and constrains the common representation matrix within the same class to be low-rank, in order to induce the intra-class representations more relevant. More formally, two regularization terms are devised for the two aspects, which have been incorporated into the objective of DSSIL. To optimize the modality-specific subnetworks and the projection layers simultaneously by exploiting the gradient decent directly, we approximate the nonconvex low-rank constraint by minimizing a few smallest singular values of the intra-class matrix with theoretical analysis. Extensive experiments conducted on three public datasets demonstrate the competitive superiority of DSSIL for cross-modal retrieval compared with the state-of-the-art methods.

Copper enrichment in the brain is highly related to Alzheimer's disease (AD) pathogenesis, but in vivo tracing of Cu2+ in the brain by imaging techniques is still a great challenge. In this work, we developed a series of activatable photoacoustic (PA) probes with low molecular weights (less than 438 Da), RPS1-RPS4, which can specifically chelate with Cu2+ to form radicals with turn-on PA signals in the near-infrared (NIR) region. Introducing the electron-donating group N,N-dimethylaniline into the probe was found to significantly enhance the radical stability and PA intensity. The best probe in the series, RPS1, showed a fast response (within seconds) to Cu2+ with high selectivity and a low PA detection limit of 90.9 nm. Owing to the low molecular weight and amphiphilic structure, RPS1 could effectively cross the blood-brain barrier (BBB) and thus allowed us, for the first time, to visualize Cu2+ in vivo via PA imaging in the brains of AD mice.

In this work, we release a multi-domain and multi-modality event dataset (MMED), containing 25,052 textual news articles collected from hundreds of news media sites (e.g., Yahoo News, BBC News, etc.) and 75,884 image posts shared on Flickr by thousands of social media users. The articles contributed by professional journalists and the images shared by amateur users are annotated according to 410 real-world events, covering emergencies, natural disasters, sports, ceremonies, elections, protests, military intervention, economic crises, etc. The MMED dataset is collected by the following the principles of high relevance in supporting the application needs, a wide range of event types, non-ambiguity of the event labels, imbalanced event clusters, and difficulty discriminating the event labels. The dataset can stimulate innovative research on related challenging problems, such as (weakly aligned) cross-modal retrieval and cross-domain event discovery, inspire visual relation mining and reasoning, etc. For comparisons, 15 baselines for two scenarios have been quantitatively and qualitatively evaluated using the dataset.

In this paper, we propose a Generative Adversarial Network with Pixel and Perceptual regularizations, denoted as P2GAN, to restore single motion blurry and low-resolution images jointly into clear and high-resolution images. It is an end-to-end neural network consisting of deblurring module and super-resolution module, which repairs degraded pixels in the motion-blur images firstly, and then outputs the deblurred images and deblurred features for further reconstruction. More specifically, the proposed P2GAN integrates pixel-wise loss in pixel-level, contextual loss and adversarial loss in perceptual level simultaneously, in order to guide on deblurring and super-resolution reconstruction of the raw images that are blurry and in low-resolution, which help obtaining realistic images. Extensive experiments conducted on a real-world dataset manifest the effectiveness of the proposed approaches, outperforming the state-of-the-art models.

In this paper, we propose a Self-attention StarGAN by introducing the self-attention mechanism into StarGAN to deal with multi-domain image-to-image translation, aiming to generate images with high-quality details and obtain consistent backgrounds. The self-attention mechanism models the long-range dependencies among the feature maps at all positions, which is not limited to the local image regions. Simultaneously, we take the advantage of batch normalization to reduce reconstruction error and generate fine-grained texture details. We adopt spectral normalization in the network to stabilize the training of Self-attention StarGAN. Both quantitative and qualitative experiments on a public dataset have been conducted. The experimental results demonstrate that the proposed model achieves lower reconstruction error and generates images in higher quality compared to StarGAN. We exploit Amazon Mechanical Turk (AMT) for perceptual evaluation, and 68.1% of all 1,000 AMT Turkers agree that the backgrounds of the images generated by Self-attention StarGAN are more consistent with the original images.

2D barcode becomes a famous technology in information security issue in this decade. We propose a security mechanism improvement method by applying the error correction with random segmentation (ECRS) technique. In this paper, we use private key array for segmentation of the original texts be-fore merging the segmented texts whose were encoded by using error correction coding approach. Our proposed method has two advantages when com-pare with the traditional Reed-Solomon error correction coding algorithm. On one hand, our ECRS performs random segmentation using multiple redundant code to protect data. On the other hand, the ECRS protects the code from being falsified by using encryption keys. Our experiments manifest that the proposed method improves the security degree of significance while just only few sacrificing coding times.

In this paper, we propose a Convolutional Neural Network(CNN) model by combining Additive Margin Softmax with Center Loss for face recognition, which is denoted as AMS-CL. AMS-CL adopts the CNN network structure of ResNet50, which is pretrained on the FaceScrub dataset. In particular, AMS-CL exploits the Additive Margin Softmax (i.e., AM-softmax) to increase inter-class distance, avoiding overconfidence on face classifications. Furthermore, a regularization term for center loss is introduced by AMS-CL to decrease the intra-class distance simultaneously. Therefore, AMS-CL keeps beneficial inter-class and intra-class relations to classification tasks jointly. Alternatively, we propose to use classifiers on the hidden data representation achieved by AMS-CL to make predictions. Extensive experiments conducted on the LFW dataset show the effectiveness of the proposed AMS-CL model.

Background Disease named entity recognition (NER) plays an important role in biomedical research. There are a significant number of challenging issues to be addressed; among these, the identification of rare diseases and complex disease names and the problem of tagging inconsistency (i.e., if an entity is tagged differently in a document) are attracting substantial research attention. Methods We propose a new neural network method named Dic-Att-BiLSTM-CRF (DABLC) for disease NER. DABLC applies an efficient exact string matching method to match disease entities with a disease dictionary; here, the dictionary is constructed based on the Disease Ontology. Furthermore, DABLC constructs a dictionary attention layer by incorporating a disease dictionary matching method and document-level attention mechanism. Finally, a bidirectional long short-term memory network and conditional random field (BiLSTM-CRF) with a dictionary attention layer is proposed to combine the disease dictionary to develop disease NER. Results Extensive experiments are conducted on two widely-used corpora: the NCBI disease corpus and the BioCreative V CDR corpus. We apply each test on 10 executions of each model, with a 95% confidence interval. DABLC achieves the highest F1 scores (NCBI: Precision = 0.883, Recall = 0.89, F1 = 0.886; BioCreative V CDR: Precision = 0.891, Recall = 0.875, F1 = 0.883), outperforming the state-of-the-art methods. Conclusion DABLC combines the advantages of both external dictionary resources and deep attention neural networks. This aids the identification of rare diseases and complex disease names; moreover, it reduces the impact of tagging inconsistency. Special disease NER and deep learning models addressing long sentences are noteworthy areas for future examination.

A deep learning-based approach to phishing detection is proposed. Specifically, websites' URLs and the characters in these URLs are mapped to documents and words, respectively, in the context of word2vec-based word embedding learning. Consequently, character embedding can be achieved from a corpus of URLs in an unsupervised manner. Furthermore, we combine character embedding with the structures of URLs to obtain the vector representations of the URLs. In particular, an URL is partitioned into the following five sections: URL protocol, sub-domain name, domain name, domain suffix, and URL path. To identify the phishing URLs, existing classification algorithms can be used smoothly on the vector representations of the URLs, avoiding laborious work on designing effective features manually and empirically. For evaluations, we collect a large-scale dataset, i.e., 1 Million Phishing Detection Dataset (1M-PD), which has been released for public use. Extensive experiments conducted on two real-world datasets show the effectiveness of the proposed approach, which achieves an accuracy of 99.69% with 0.40% false positive and 99.79% true positives on the 1M-PD dataset. In particular, the proposed approach detects each URL in 32ms on average merely on a personal computer, which is much faster than existing approaches and even can be considered real-time.

Deduplication scheme based on convergent encryption (CE) is widely-used in cloud storage system to eliminate redundant data. However, the adversaries can obtain the data by the brute-force attack, if the data belongs to a predictable set for CE. In addition, previous works usually introduce the third-party auditors to execute the data integrity verification, suffering from data disclosure by the auditors. In this paper, we propose a secure authorized deduplication scheme based on blockchain, which can ensure the confidentiality and security of the users' data stored on cloud servers. In our scheme, the users can utilize the smart contract to create a tamper-proofing ledger, which can protect the data from illegal modification. Meanwhile, the users can execute the integrity audit protocols to check the users’ data integrity by the smart contract. Moreover, a hierarchical role hash tree (HRHT) will be constructed to create the role key when the users upload their data to CSP, allowing the authorized users to access the data. Security analysis and performance evaluation demonstrate that our proposed scheme is resilient against the brute-force attack and the collusion attack, while it has limited computation overhead.

Phenolic impregnated graphite is a new kind of material with higher corrosion resistance and thermal conductivity. In the chemical industry, it is often used to make containers and pipes containing strong acid and alkali medium. In this paper, the working conditions of impregnated graphite materials in the actual production were simulated in the laboratory, and the. mechanical properties and material characterization of impregnated graphite after soaking in sulfuric acid at different temperatures were compared. The results show that the tensile strength of phenolic resin impregnated graphite decreases with the increase of immersion temperature under the action of acid medium and temperature; the results of thermogravimetric analysis show that the inner phenolic resin expands and extrudes graphite particles to produce internal stress, and the graphite particles will fall off partially, but the chemical structure of phenolic resin does not change.

Phenolic impregnated graphite is a new kind of material with higher corrosion resistance and thermal conductivity. In the chemical industry, it is often used to make containers and pipes containing strong acid and alkali medium. It is the characteristic of phenolic resin impregnated graphite heat exchanger that the corrosion resistance and heat transfer efficiency of the heat exchanger are greatly improved. In this paper, a failure case of graphite heat exchanger in a chemical enterprise in Shanghai was analyzed by numerical simulation. According to the actual working conditions, the temperature field and flow field in the porous graphite heat exchanger were numerically simulated. The numerical simulation results of temperature field showed that the graphite block would be subject to thermal shock when the equipment was started and shut down, resulting in the increase of temperature difference of graphite block In addition, the pre tightening force of the flange of graphite heat exchanger was simulated by numerical simulation, and it was found that the position of stress concentration coincided with the crack position of actual failure, and the improvement suggestions were put forward for the structural design of graphite block.

Phenolic impregnated graphite is a new kind of material with higher corrosion resistance and thermal conductivity. In the chemical industry, it is often used to make containers and pipes containing strong acid and alkali medium. In this paper, a failure case of graphite heat exchanger in a chemical enterprise in Shanghai was studied. The failure reason of the graphite block was tested by means of three-dimensional observation, thermogravimetry and scanning electron microscopy. The results show that the particle fineness of graphite block and the operating conditions of the graphite heat exchanger are the main reasons for the failure of the graphite heat exchanger. At last, some suggestions are put forward for the safe use of this kind of graphite heat exchanger.

In this paper, we focus on event discovery by utilizing data distributed in multiple media domains, such as news media and social media. To this end, we propose an in-domain and cross-domain Laplacian regularization (ICLR) model, which can learn effective data representation for both textual news reports contributed by journalists in news media domain, and image posts shared by amateur users in social media domain. The achieved data representation can be used by classification and clustering strategies for existing and new event discovery, respectively. More specifically, ICLR constructs respective Laplacian regularization terms considering the property of inter-domain and intra-domain label consistency, which can be optimized by employing an alternating optimization strategy with theoretical guarantee for convergence. In particular, we collect and release a multi-domain and multimodal dataset for evaluations and public use.

Circular RNAs are a class of non-coding RNAs that are receiving extensive attention. Despite reports showing circular RNAs acting as microRNA sponges, the biological functions of circular RNAs remain largely unknown. We show that in patient tumor samples and in a panel of cancer cells, circ-Foxo3 was minimally expressed. Interestingly, during cancer cell apoptosis, the expression of circ-Foxo3 was found to be significantly increased. We found that silencing endogenous circ-Foxo3 enhanced cell viability, whereas ectopic expression of circ-Foxo3 triggered stress-induced apoptosis and inhibited the growth of tumor xenografts. Also, expression of circ-Foxo3 increased Foxo3 protein levels but repressed p53 levels. By binding to both, circ-Foxo3 promoted MDM2-induced p53 ubiquitination and subsequent degradation, resulting in an overall decrease of p53. With low binding affinity to Foxo3 protein, circ-Foxo3 prevented MDM2 from inducing Foxo3 ubiquitination and degradation, resulting in increased levels of Foxo3 protein. As a result, cell apoptosis was induced by upregulation of the Foxo3 downstream target PUMA.

In this work, we aim to discover real-world events from Flickr data by devising a three-stage event detection framework. In the first stage, a multimodal fusion (MF) model is designed to deal with the heterogeneous feature modalities possessed by the user-shared data, which is advantageous in computation complexity. In the second stage, a dual graph regularized non-negative matrix factorization (DGNMF) model is proposed to learn compact feature representations. DGNMF incorporates Laplacian regularization terms for the data graph and base graph into the objective, keeping the geometry structures underlying the data samples and dictionary bases simultaneously. In the third stage, hybrid clustering algorithms are applied seamlessly to discover event clusters. Extensive experiments conducted on the real-world dataset reveal the MF-DGNMF-based approaches outperform the baselines.

The vertebrate blood-brain barrier (BBB) creates an obstacle for central nervous system-related drug delivery. Claudin-5 (Cldn5), expressed in large quantities in BBB, plays a vital role in restricting BBB permeability. The C-terminal domain of Clostridium perfringens enterotoxin (cCPE) has been verified as binding to a subset of claudins (Cldns). The Cldn5-binding cCPE194-319 variant cCPEY306W/S313H was applied in this study to investigate its ability to modulate the permeability of zebrafish larval BBB. In vitro results showed that cCPEY306W/S313H is able to bind specifically to Cldn5 in murine brain vascular endothelial (bEnd.3) cells, and is transported along with Cldn5 from the cell membrane to the cytoplasm, which in turn results in a reduction in transendothelial electrical resistance (TEER). Conversely, this effect can be reversed by removal of cCPEY306W/S313H. In an in vivo experiment, this study estimates the capability of cCPEY306W/S313H to modulate Cldn5 using a rhodamine B-Dextran dye diffusion assay in zebrafish larval BBB. The results show that cCPEY306W/S313H co-localized with Cldn5 in zebrafish cerebral vascular cells and modulated BBB permeability, resulting in dye leakage. Taken together, this study suggests that cCPEY306W/S313H has the capability - both in vitro and in vivo - to modulate BBB permeability temporarily by specific binding to Cldn5.

Full understanding on the characteristic and behavior of manufactured sand (MS) is very crucial to its application. The particle shape, surface texture and behavior of MS are characterized in this paper. Results indicate that MS has higher roundness and Length-width ratio, and wider distribution ranges of those parameters compared with river sand (RS), in the micro scale, most MS has lower surface roughness than RS unexpectedly. To obtain the similar workability, most MS concretes require higher water reducer dosage than RS concrete, and MSs with less stone powder and clay lump content require even lower water reducer. Nearly all MS concrete has higher strength than RS concrete with same paste composition. The particle shape and surface texture of MS has less significant effect on its behavior in concrete than the stone powder, clay lump content and the gradation of MS, so MS with suitable production process may have better behavior in concrete than RS.

Carbon/carbon composites, solely composed of pure carbon, are typical fiber fabric reinforced porous composites. They have been widely studied due to their excellent bio-inert performance and mechanical properties. However, their applications are restricted by two problems, i.e., easy dissociation of surface carbon particles and poor wettability of common bioactive coatings on their surface. To overcome these problems, we have adopted a biomimetic mineralization method, where Mg-doped hydroxyapatite particles were successfully prepared on hydroxyl-modified C/C surfaces, prepared via HNO3 oxidation. In this way, the original porous morphology of C/C, which is conducive to cell proliferation, is well-preserved. In the meanwhile, the introduction of electrochemical methods sped up the process of bio-mineralization and facilitated the construction of bone-like surface layers. In vivo tests revealed that 90 days after implantation for bone defect repair, the resulting C/C displayed unprecedented surface chemical stability and biocompatibility.

The purpose of this experiment was to evaluate the effects of low protein corn-soybean meal-based diets on fecal CP, amino acid (AA) flow amount, AA digestibility and fecal and ileal microbial AA composition in growing pigs. Eighteen pigs (initial body weight = 30 &plusmn, 1.35) were randomly divided into three groups and fed with basal diets with CP levels of 12%, 15% and 18%, respectively. The Lys, Met + Cys, Thr and Trp level in the 12% CP and 15% CP groups is the same as 18% CP group by the addition of four crystalline Lys, Met + Cys, Thr and Trp to the diet. The results showed that with the decrease of dietary CP level from 18% to 12%, the fecal total nitrogen (N), CP and total AA (TAA) flow amount decreased linearly (p &lt, 0.05). Dry matter (DM) digestibility, CP digestibility, TAA digestibility, essential amino acid (EAA) digestibility and non-essential amino acid (NEAA) digestibility increased linearly with the decrease of dietary CP concentration from 18% to 12%. Compared with 18% CP group, the flow amount of Asp, Ser, Glu, Gly, Tyr, Val, Leu and Phe in feces of pigs in the 15% CP group and 12% CP group decreased significantly, while the flow amount of Arg in the 15% CP group was lower than that in the 18% CP group and 12% CP group. The fecal microbial N and AA of the 15% CP group were higher than those of the 18% CP and 12% CP groups. Fecal TAA flow amount decreased linearly with the decrease of the dietary CP levels from 18% to 12%. Fecal TAA and NEAA flow amount also decreased linearly with the decrease of dietary CP level from 18% to 12%. Except for Glu, Gly, Met, Tyr, Thr and Phe, there were significant differences among the three groups in the composition of 17 kinds of AAs in fecal microorganisms. Among the 17 AA compositions of ileal microorganisms, except Tyr and Lys, the other AAs were significantly different among the three groups (p &lt, 0.05)

A zeolite supported organic corrosion inhibitor with high loading efficiency was obtained by loading a synthesized lauric acid imidazoline (LAI) in the Ca2+-exchanged zeolite (Ca2+-zeolite) framework. The influences of the as-prepared Ca2+-zeolite supported LAI (Ca2+-ZSLAI) as an admixture in porosity and compressive strength of the resulting cement mortar were examined. The protective effect of Ca2+-ZSLAI against steel corrosion of steel-mortar immersed in NaCl solution was discussed upon. The properties of the steel reinforced mortar incorporating Ca2+-ZSLAI was compared with the reference mortar specimen without corrosion inhibitor and the specimen solely with LAI or Ca2+-zeolite. The steel corrosion resistance performance of Ca2+-ZSLAI was revealed by electrochemical testing on steel reinforced mortar during the chloride ions accelerated corrosion test. Experimental results revealed that mortar with 6% of cement replaced by Ca2+-zeolite saturatedly supported LAI as corrosion inhibition additive presented 14% higher compressive strength and 23% lower total porosity compared to the reference mortar. Besides, such dose of inhibitor could retain 92% corrosion inhibition efficiency of the embedded steel after 142 immersion days. The excellent corrosion resistance efficacy of Ca2+-ZSLAI was ascribed to the chloride “binding capacity” of the densified mortar substrate improved by the filling and pozzolanic effect of Ca2+-zeolite and the chloride “barrier capacity” of the protective interfacial film deposited by the LAI released from the Ca2+-zeolite carrier.

Scholars' homepages are important places to show personal research interest and academic achievement through the Web. However, according to our observation, only a small portion of scholars update their publications and related events on their homepages in time. In this paper, we propose a homepage augmentation technique, which automatically shows the newest academic events related to a scholar on his/her homepage. Specifically, we model the relations between homepages and the events collected from the Web as a complex heterogenous network, and propose an Embedding-based Heterogenous random Walk algorithm, namely EHWalk, to predict the links between homepages and events. Compared with existing embedding-based link prediction algorithms, EHWalk supports more efficient modeling of complex heterogenous relations in a dynamically changing network, which helps link the massive new updated events to homepages precisely and efficiently. Comprehensive experiments on a real-world dataset are conducted and the results show that our algorithm can achieve both good effectiveness and efficiency for real-world deployment.

Cross-domain beauty and personal care product image retrieval is a challenging problem due to data variations (e.g., brightness, viewpoint, and scale), and the rich types of items. In this paper, we present a regional maximum activations of convolutions with attention (RA-MAC) descriptor to extract image features for retrieval. RA-MAC improves the regional maximum activations of convolutions (R-MAC) descriptor considering the influence of background in cross-domain images (i.e., shopper domain and seller domain). More specifically, RA-MAC utilizes the characteristics of the convolutional layer to find the attention of an image, and reduces the influence of the unimportant regions in an unsupervised manner. Furthermore, a few strategies have been exploited to improve the performance, such as multiple features fusion, query expansion, and database augmentation. Extensive experiments conducted on a dataset consisting of half a million images of beauty care products (Perfect-500K) manifest the effectiveness of RA-MAC. Our approach achieves the 2nd place in the leader board of the Grand Challenge of AI Meets Beauty in ACM Multimedia 2018. Our code is available at: https://github.com/RetrainIt/Perfect-Half-Million-Beauty-Product-Image-Recognition-Challenge.

Social media headline prediction (SMHP) is a thriving application scenario, which aims to predict the popularity of the post data shared on social media. In this paper, we propose to use multi-aspect features combined with the random forest (RF) model for popularity predictions. Firstly, we extract features by combining both metadata of the posts and users' features. More specifically, we adopt the binary coding strategy for dimensionality reduction and deal with the missing values by using some strategies, i.e., estimating the missing geographic information according to the information of users, and filling the missing features with median. Furthermore, regression models can be used directly to make predictions. In particular, a random forest (RF) model is adopted since it does not require much effort in tuning hyper-parameters and performs effectively. Extensive experiments conducted on the SMHP dataset consisting of 340K image posts shared by 80K users manifest the effectiveness of our method. Our approach achieves the 4nd place in the leader board of the Grand Challenge of SMHP in ACM Multimedia 2018.

An online community based on blockchain is proposed for software developers to share, assess, and learn codes and other codes or software related knowledge. It involves three modules or roles, namely: developer (or coder, or more generally, knowledge contributor), code (or knowledge contribution), and jury (or assessor, who is usually a developer with advanced skills), in addition to the blockchain based database. Each full node of the blockchain hosts a copy of all activities of developers in such community, including uploading contributions, assessing others’ contributions, and conducting transactions. Smart contracts are applicable to automate transactions after code assessment or other related activities. The system aims to assess and improve the value of codes accurately, stimulate the creativity of the developers, and improve software development efficiency, so as to establish a virtuous cycle of a software development community.

An insurance system based on blockchain is proposed for web identity security, which provides two insurance service models for personal web identity security of end users and data security of commercial websites, respectively. Claim evidences are uploaded automatically to the blockchain to keep their authenticity. Smart contracts are automatically applied between insurers and policyholders to build trust between them.

It has been demonstrated that the deregulation of microRNAs (miRNAs) affects the development of rheumatoid arthritis (RA). The primary objective of the current study was to determine the role of miR-143-3p in the progression of RA. The expression of miR-143-3p in synovium taken from patients with RA was assessed by reverse transcription-quantitative polymerase chain reaction. The expression of miR-143-3p was higher in synovium tissues of RA than that of osteoarthritis (OA). The decreased expression of miR-143-3p suppressed cell proliferation and promoted apoptosis in vitro. In addition, inhibition of miR-143-3p decreased levels of inflammatory cytokines, as determined by an enzyme-linked immunosorbent assay. IGF1R and IGFBP5 were found to be the target genes of miR-143-3p, and it was demonstrated that miR-143-3p regulated the proliferation and apoptosis of MH7A cells by targeting IGF1R and IGFBP5. Furthermore, TNF-α treatment stimulated the Ras/p38 mitogen activated protein kinase (MAPK) signaling pathway, whereas miR-143-3p inhibition suppressed it. The results of the current study indicate that miR-143-3p may regulate cell proliferation and apoptosis by targeting IGF1R and IGFBP5 expression and regulating the Ras/p38 MAPK signaling pathways. Therefore, miR-143-3p may be a novel therapeutic target in RA.

Circular RNAs are a large group of noncoding RNAs that are widely expressed in mammalian cells. Genome-wide analyses have revealed abundant and evolutionarily conserved circular RNAs across species, which suggest specific physiological roles of these species. Using a microarray approach, we detected increased expression of a circular RNA circ-Dnmt1 in eight breast cancer cell lines and in patients with breast carcinoma. Silencing circ-Dnmt1 inhibited cell proliferation and survival. Ectopic circ-Dnmt1 increased the proliferative and survival capacities of breast cancer cells by stimulating cellular autophagy. We found that circ-Dnmt1-mediated autophagy was essential in inhibiting cellular senescence and increasing tumor xenograft growth. We further found that ectopically expressed circ-Dnmt1 could interact with both p53 and AUF1, promoting the nuclear translocation of both proteins. Nuclear translocation of p53 induced cellular autophagy while AUF1 nuclear translocation reduced Dnmt1 mRNA instability, resulting in increased Dnmt1 translation. From here, functional Dnmt1 could then translocate into the nucleus, inhibiting p53 transcription. Computational algorithms revealed that both p53 and AUF1 could bind to different regions of circ-Dnmt1 RNA. Our results showed that the highly expressed circular RNA circ-Dnmt1 could bind to and regulate oncogenic proteins in breast cancer cells. Thus circ-Dnmt1 appears to be an oncogenic circular RNA with potential for further preclinical research.

In this paper, we propose to learn Deep Semantic Space (DSS) for cross-modal event retrieval, which is achieved by exploiting deep learning models to extract semantic features from images and textual articles jointly. More specifically, a VGG network is used to transfer deep semantic knowledge from a large-scale image dataset to the target image dataset. Simultaneously, a fully-connected network is designed to model semantic representation from textual features (e.g., TF-IDF, LDA). Furthermore, the obtained deep semantic representations for image and text can be mapped into a high-level semantic space, in which the distance between data samples can be measured straightforwardly for cross-model event retrieval. In particular, we collect a dataset called Wiki-Flickr event dataset for cross-modal event retrieval, where the data are weakly aligned unlike image-text pairs in the existing cross-modal retrieval datasets. Extensive experiments conducted on both the Pascal Sentence dataset and our Wiki-Flickr event dataset show that our DSS outperforms the state-of-the-art approaches.

Recently, domain adaptation has gained great popularity, while most researchers are focusing on domains in homogenous modalities, e.g., image domains. In reality, heterogeneous domains are pretty common and more challenging. In this paper, we present MCD-JD—a Maximum Classifier Discrepancy model which considers the joint distribution of the source and target domain data for heterogeneous domain adaption. MCD-JD derives from Generative Adversarial Networks (GAN) consisting of two parts, i.e., minimizing the discrepancy of joint distribution, and maximizing classifier discrepancy. Specifically, the first part uses the Maximum Mean Discrepancy (MMD) regularization to adapt the data distributions between source and target domains. The second part utilizes two different classifiers to maximize their discrepancy of making predictions on the target domain data, which further minimizes the discrepancy of data distributions between source and target domains. We collect a dataset depicting real-world events (e.g., protests, explosions, etc.) from multiple heterogeneous data domains, including news media textual articles, social media (Flickr) images, and YouTube videos. Extensive experiments conducted on the real-world dataset manifest the effectiveness of MCD-JD, which outperforms state-of-the-art benchmark models.

Ferritic/martensitic (F/M) steels are primary candidates for application as cladding and structural materials in the Generation IV Nuclear Reactor, especially accelerator driven sub-critical system (ADS). The compatibility of F/M steels with liquid lead-bismuth eutectic (LBE) is a critical issue for development of ADS using liquid LBE as the coolant. In this work, the corrosion tests of two F/M steels, including a novel 9–12 Cr modified F/M steel named SIMP steel and a commercial T91 steel, were conducted in the static oxygen-saturated liquid LBE at 600 °C up to 1000 h, the microstructure of the oxide scale formed on these two steels was analyzed, the relationship between the microstructure and the oxidation behavior was studied, and the reason why the SIMP steel showed better oxidation resistance compared to T91 steel was analyzed. The results of this study confirmed that the oxidation behavior of the F/M steels in liquid metals is influenced by their alloying elements and microstructures.

The effect of electromagnetic stirring on molten steel flow and heat transfer in a 260 mm×300 mm bloom mold was investigated by using a method combining both finite element method and finite volume method. The simulation results related to magnetic fields were consistent with the onsite measured data. The magnetic flux density increased with increasing the current intensity but decreased with increasing the current frequency. Electromagnetic stirring caused molten steel to flow with rotation on a horizontal section and two sets of recirculation regions with opposite reflow directions in a longitudinal section formed. The maximum tangential velocity increased with increasing the current intensity and frequency. Furthermore, the superheat degree of the molten steel on the outlet cross section of the mold decreased with increasing the current intensity. Electromagnetic stirring caused the emergence of a zero growth zone of solidified shell in the effective stirring zone. For the 260 mm×300 mm bloom continuous caster of bearing steel, the appropriate values of current intensity and current frequency of electromagnetic stirring were found to be 300 A and 3 Hz, respectively.

A hybrid model for social media popularity prediction is proposed by combining Convolutional Neural Network (CNN) with XGBoost. The CNN model is exploited to learn high-level representations from the social cues of the data. These high-level representations are used in XGBoost to predict the popularity of the social posts. We evaluate our approach on a real-world Social Media Prediction (SMP) dataset, which consists of 432K Flickr images. The experimental results show that the proposed approach is effective, achieving the following performance: Spearman's Rho: 0.7406, MSE: 2.7293, MAE: 1.2475.

Background Intra-uterine growth restriction (IUGR) and fetal overgrowth increase risks to postnatal health. Maternal nutrition is the major intrauterine environmental factor that alters fetal weight. However, the mechanisms underlying the effects of maternal nutrition on fetal development are not entirely clear. We developed a pig model, and using isobaric tags for relative and absolute quantification (iTRAQ), we investigated alterations in the placental proteome of gilts on a normal-energy-intake (Con) and high-energy-intake (HE) diet. Results In the Con group, heavy and light fetuses were found at the tubal and cervical ends of the uterus respectively at 90 d of gestation. Moreover, the heavy fetuses had a higher glucose concentration than the light fetuses. However, a higher uniformity was noted in the HE group. Placental promoters between these two positions indicated that 78 and 50 differentially expressed proteins were detected in the Con and HE groups respectively. In the Con group, these proteins were involved in lipid metabolism (HADHA, AACS, CAD), nutrient transport (GLUT, SLC27A1), and energy metabolism (NDUFV1, NDUFV2, ATP5C1). However, in the HE group they mainly participated in transcriptional and translational regulation, and intracellular vesicular transport. Conclusions Our findings revealed that maternal nutrition may alter birth weight mainly through the modulation of placental lipid and energy metabolism, which also provides a possible mechanism to explain the higher uniformity of fetal weight in gilts fed a HE diet. Electronic supplementary material The online version of this article (doi:10.1186/s12864-017-3601-1) contains supplementary material, which is available to authorized users.

Online news media and social media are popular domains for people to acquire real-world event knowledge. In this work, the problem of multi-domain and multi-modality event detection (MMED) is elaborated. We wish to organize the multi-modality data from multiple domains based on real-world events. To this end, a cross-domain and cross-modality transfer learning (CDM) model is proposed. The CDM model aligns the data by exploiting a dictionary-based alignment strategy, and identifies the event labels of the data samples based on the class-specific reconstruction residual. Extensive experiments conducted on real-world data demonstrate the effectiveness of the proposed models. In particular, a benchmark dataset, denoted as MMED100, is released, which can hopefully be used to promote the research on this topic and advance related applications.

The publicly available data such as the massive and dynamically updated news and social media data streams (a.k.a. big data) covers the various aspects of social activities, personal views and expressions, which points to the importance of understanding and discovering the knowledge patterns underlying the big data, and the need of developing methodologies and techniques to discover real-world events from such big data, to manage and to analyze the discovered events in an effective and elegant way. In this paper we present an event cube (EC) model which is devised to support various queries and analysis tasks of events; such events include those discovered by techniques of untargeted event detection (UED) and targeted event detection (TED) from multi-sourced data. Specifically, based on the essential event elements of 5W1H (i.e., When, Where, Who, What, Why, and How), the EC model is developed to organize the discovered events from multiple dimensions, to operate on the events at various levels of granularity, so as to facilitate analyzing and mining hidden/inherent relationships among the events effectively. Case studies are provided to illustrate the usages and show the benefits of EC facilities in on-line analytical processing of events and their relationships.

A new type of concrete named "safety concrete" for traffic safety is developed. A no-fine concrete solidified by Alkali Activated Slag (AAS) can form a network of micro-cracks throughout the concrete to ensure its good energy absorption under the sudden and severe loading, the mixture design and microstructure of this new concrete is studied. RESULTS indicated that, the optimum aggregate gradation with 2.36 mm passing ratio around 20% and around 5% of particles bigger than 4.75 mm ensure this concrete reach compressive strength of 20 MPa and fragmentability of 90%, the micro-cracks generate in different scales in the paste of this concrete, there are much more micro-cracks generated in AAS than in OPC paste, laboratory manufactured no-fine safety concrete blocks are broken into fragments smaller than 4.75 mm when they are dropped from 7th floor (18 m), safety concrete can serve as frontier guard railing or crash barrier installed for highway traffic safety. 2014 Elsevier Ltd. All rights reserved. Language: en

Within-litter uniformity in pigs is a major factor affecting piglet survival and growth performance. We know that Meishan (MS) gilts have higher piglet survival rate than Large White (LW) gilts because their foetal weight is less varied. To understand the molecular basis for placental nutritional transport during the late stages of gestation in LW and MS, we employed the isobaric tags for relative and absolute quantification (iTRAQ) method to investigate alterations in the placental proteomes of LW and MS gilts on gestational day 90. Investigation of foetal weight at different uterine positions revealed that the foetal and placental weights as well as the foetal concentration of glucose were significantly higher in LW gilts positioned towards the utero-tubal junction than in those positioned toward the cervix; however, no such differences were observed in MS gilts, and MS gilts had a greater uniformity in foetal weight on day 90 of gestation. Comparisons of the proteomes between placentas positioned toward the cervix and those positioned toward the utero-tubal junction identified 38 differentially expressed proteins in the two breeds. These proteins play a central role in nutrient transport and metabolism, as well as in transcriptional and translational regulation. Of particular interest is the finding that the placentas of LW gilts showed 14 differential expression of proteins mainly related to lipid transport and energy metabolism (including solute carrier family 27, mitochondrial trifunctional protein, and NADH dehydrogenase [ubiquinone] flavoprotein 2), but only 2 proteins in MS gilts. In contrast, the differentially expressed proteins in MS gilts were primarily involved in transcriptional and translational regulation (such as ribosome-sec61 and 40S ribosomal protein S23), with a few related to glucose and coenzyme transport and metabolism (including glucose transport protein and ferrochelatase). Our results revealed that placental lipid and energy metabolism might play a crucial role in the regulation of foetal weight, based on uterine position in two distinct pig breeds. These findings provide a deeper understanding of placental efficiency that can be utilized to provide a new method to enhance the efficiency of livestock production.

Internet platforms provide new ways for people to share experiences, generating massive amounts of data related to various real-world concepts. In this paper, we present an event detection framework to discover real-world events from multiple data domains, including online news media and social media. As multi-domain data possess multiple data views that are heterogeneous, initial dictionaries consisting of labeled data samples are exploited to align the multi-view data. Furthermore, a shared multi-view data representation (SMDR) model is devised, which learns underlying and intrinsic structures shared among the data views by considering the structures underlying the data, data variations, and informativeness of dictionaries. SMDR incorpvarious constraints in the objective function, including shared representation, low-rank, local invariance, reconstruction error, and dictionary independence constraints. Given the data representations achieved by SMDR, class-wise residual models are designed to discover the events underlying the data based on the reconstruction residuals. Extensive experiments conducted on two real-world event detection datasets, i.e., Multi-domain and Multi-modality Event Detection dataset, and MediaEval Social Event Detection 2014 dataset, indicating the effectiveness of the proposed approaches.

Between January 2012 and March 2012, the infection rates of porcine epidemic diarrhea virus (PEDV) increased substantially in vaccinated swine herds in many porcine farms in Gansu Province, China. The spike (S) glycoprotein is an important determinant for PEDV biological properties. To determine the distribution profile of PEDV outbreak strains, we sequenced the full-length S gene of five samples from two farms where animals exhibited severe diarrhea and high mortality rates. Five new PEDV variants were identified, and the molecular diversity, phylogenetic relationships, and antigenicity analysis of Gansu field samples with other PEDV reference strains were investigated. A series of insertions, deletions, and mutations in the S gene was found in five PEDV variants compared with classical and vaccine strains. These mutations may provide stronger pathogenicity and antigenicity to the new PEDV variants that influenced the effectiveness of the CV777-based vaccine. Our results suggest that these new PEDV variant strains in Gansu Province might be from South Korean or South China, and the effectiveness of the CV777-based vaccine needs to be evaluated.

Long-term electrical resistance tests were performed to evaluate the performance of AISI 441 ferritic stainless steel coated with a Mn–Co spinel protection layer as a candidate material for intermediate temperature solid oxide fuel cell interconnect applications. The tests indicated that, while uncoated AISI 441 shows a substantial increase in area-specific electrical resistance (ASR), spinel-coated AISI 441 exhibits much lower ASR values. The spinel coatings reduced the oxide scale growth rate and blocked outward diffusion of Cr from the oxide scale that grew between the coating and the steel substrate during the long-term tests. The oxide scale consisted of a chromia layer containing discrete regions of Mn–Cr spinel distributed throughout the layer. The presence of Ti in the chromia scale matrix and/or the presence of regions of Mn–Cr spinel within the scale may have increased the scale electrical conductivity, which would explain the fact that the observed ASR in the tests was lower than would be expected if the scale consisted of pure chromia.

The recently invented iron (Fe)/vanadium (V) redox flow battery (IVB) system has attracted increasing attention because of its long-term cycling stability and low-cost membrane/separator. In this paper, we describe our extensive matrix study of factors such as electrolyte composition, state of charge (SOC), and temperature that influence the stability of electrolytes in both positive and negative half-cells. During the study, an optimized electrolyte that can be operated in a temperature range from −5 °C to 50 °C without precipitation is identified. Fe/V flow cells using the optimized electrolyte and low-cost separator exhibit satisfactory cycling performance at different temperatures. Efficiencies, capacities, and energy densities of flow batteries at various temperatures are studied.

High temperature (250–350 °C) sodium–beta alumina batteries (NBBs) are attractive energy storage devices for renewable energy integration and other grid related applications. Currently, two technologies are commercially available in NBBs, e.g., sodium–sulfur (Na–S) battery and sodium–metal halide (ZEBRA) batteries. In this study, we investigated the combination of these two chemistries with a mixed cathode. In particular, the cathode consisted of molten NaAlCl4 as a catholyte and a mixture of Ni, NaCl and Na2S as active materials. During cycling, two reversible plateaus were observed in cell voltage profiles, which matched electrochemical reactions for Na–S and Na–NiCl2 redox couples. An irreversible reaction between sulfur species and Ni was identified during initial charge at 280 °C, which caused a decrease in cell capacity. The final products on discharge included Na2Sn with 1

In this study, we reported an intermediate-temperature (∼150 °C) sodium–sulfur (Na–S) battery. With a relatively low operating temperature, this novel battery could reduce the cost and safety issues associated with the conventional high-temperature (300–350 °C) Na–S battery. A dense β′′-Al2O3 solid membrane and tetraglyme were utilized as the electrolyte separator and catholyte solvent in this battery. Solubility tests indicated that a cathode mixture of Na2S4 and S exhibited extremely high solubility in tetraglyme (e.g., >4.1 M for Na2S4 + 4 S). CV scans of Na2S4 in tetraglyme revealed two pairs of redox couples with peaks at around 2.22 and 1.75 V, corresponding to the redox reactions of polysulfide species. The discharge/charge profiles of the Na–S battery showed a slope region and a plateau, indicating multiple steps and cell reactions. In situ Raman measurements during battery operation suggested that polysulfide species were formed in the sequence of Na2S5 + S → Na2S5 + Na2S4 → Na2S4 + Na2S2 during discharge and in a reverse order during charge. This battery showed dramatic improvement in rate capacity and cycling stability over room-temperature Na–S batteries, which makes it more attractive for renewable energy integration and other grid related applications.

The sodium–nickel chloride (ZEBRA) battery is operated at relatively high temperature (250–350 °C) to achieve adequate electrochemical performance. Reducing the operating temperature in the range of 150200 °C can not only lead to enhanced cycle life by suppressing temperature-related degradations, but also allow the use of lower cost materials for construction. To achieve adequate electrochemical performance at lower operating temperatures, reduction in ohmic losses is required, including the reduced ohmic resistance of β″-alumina solid electrolyte (BASE) and the incorporation of low melting point secondary electrolytes. In present work, planar-type Na/NiCl 2 cells with a thin BASE (600 μm) and low melting point secondary electrolyte were evaluated at reduced temperatures. Molten salts used as secondary electrolytes were fabricated by the partial replacement of NaCl in the standard secondary electrolyte (NaAlCl 4 ) with other lower melting point alkali metal salts such as NaBr, LiCl, and LiBr. Electrochemical characterization of these ternary molten salts demonstrated improved ionic conductivity and sufficient electrochemical window at reduced temperatures. Furthermore, Na/NiCl 2 cells with 50 mol% NaBr-containing secondary electrolyte exhibited reduced polarizations at 175 °C compared to the cell with the standard NaAlCl 4 catholyte. The cells also exhibited stable cycling performance even at 150 °C.