Search

In 1-bit matrix completion, the aim is to estimate an underlying low-rank matrix from a partial set of binary observations. We propose a novel method for 1-bit matrix completion called MMGN. Our method is based on the majorization-minimization (MM) principle, which yields a sequence of standard low-rank matrix completion problems in our setting. We solve each of these sub-problems by a factorization approach that explicitly enforces the assumed low-rank structure and then apply a Gauss-Newton method. Our numerical studies and application to a real-data example illustrate that MMGN outputs comparable if not more accurate estimates, is often significantly faster, and is less sensitive to the spikiness of the underlying matrix than existing methods., Comment: 33 pages, 9 figures

G protein–coupled receptors can signal through both G proteins and ß-arrestin2. For the µ-opioid receptor (MOR), early experimental evidence from a single study suggested that G protein signaling mediates analgesia and sedation, whereas ß-arrestin signaling mediates respiratory depression and constipation. Then, receptor mutations were used to clarify which residues interact with ligands to selectively regulate signals in a ligand-specific manner. However, there is no systematic study on how to determine these residues and clarify the molecular mechanism of their influence on signal pathways. We have therefore used molecular docking to predict the amino acid sites that affect the binding of ligands and MOR. Then, the corresponding sites were mutated to determine the effect of the structural determinant of MOR on Gi/o protein and ß-arrestin pathways. The pharmacological and animal behavioral experiments in combination with molecular dynamics simulations were used to elucidate the molecular mechanism of key residues governing the signaling. Without affecting ligand binding to MOR, MORY7.43A attenuated the activation of both Gi/o protein and ß-arrestin signaling pathways stimulated by fentanyl, whereas it did not change these two pathways stimulated by morphine. Likewise, the activation peak time of extracellular regulated protein kinases was significantly prolonged at MORY7.43A compared with that at MORwildtype stimulated by fentanyl, but there was no difference stimulated by morphine. In addition, MORY7.43A significantly enhanced analgesia by fentanyl but not by morphine in the mice behavioral experiment. Furthermore, the molecular dynamics simulations showed that H6 moves toward the cellular membrane. H6 of the fentanyl–Y7.43A system moved outward more than that in the morphine–Y7.43A system. Y7.43 mutation disrupted hydrophobic interactions between W6.48 and Y7.43 in the fentanyl–Y7.43A system but not in the morphine–Y7.43A system. Our results have disclosed novel mechanisms of Y7.43 mutation affecting MOR signaling pathways. Y7.43 mutation reduced the activation of the Gi/o protein pathway and blocked the ß-arrestin2 recruitment, increased the H6 outward movement of MOR, and disrupted hydrophobic interactions. This may be responsible for the enhanced fentanyl analgesia. These findings are conducive to designing new drugs from the perspective of ligand and receptor binding, and Y7.43 is also expected to be a key site to structure optimization of synthesized compounds.

Background: In this study, modulators of human Chemotactic cytokine receptor 5 (CCR5) were described using a quantitative structure-activity relationship model (QSAR). This model was based on the molecule’s chemical structure. Methods:: All 56 compounds of CCR5 receptor antagonists were randomly separated into two sets, 43 were reserved for training and the other 13 for testing. In the course of this study, molecular models were drawn using ChemDraw software. By means of Hyperchem software as well as optimized with both AM1 (semi-empirical self-consistent-field molecular orbital) and MM+ (molecular mechanics plus force field), molecular models were described through numerous descriptors using CODESSA software. Results: Linear models were obtained by Heuristic Method (HM) software and nonlinear models were obtained using APS software with optimal descriptor combinations used to build linear QSAR models, involving a group of selected descriptors. As a result, values of the above two different sets were shown to result from 0.82 in testing and 0.86 in training in HM while 0.83 in testing and 0.88 in training in Gene Expression Programming (GEP). Conclusion: From this method, the activity of molecules could be predicted, and the molecular structure could be changed to alter its IC50, avoiding the testing of large numbers of compounds.

Pattern recognition has made great progress under large amount of labeled data, while performs poorly on a very few examples obtained, named few-shot classification, where a classifier can identify new classes not encountered during training. In this paper, a simple framework named Prototype-Relation Network is presented for the few-shot classification. Moreover, a novel loss function compared with prototype networks is proposed which takes both inter-class and intra-class distance into account. During meta-learning, the model is optimized by end-to-end episodes, each of which is to imitate the test few-shot setting. The trained model is used to classify new classes by computing min distance between query images and the prototype of each class. Extensive experimental results demonstrate that our proposed meta-learning model is competitive and effective, which achieves the state-of-the-art performance on Omniglot and miniImageNet datasets.

Personality affects individual’s work and life style, and has important guiding significance for individual’s psychological counseling and career development. Traditionally, researchers use scales to measure personality. However, in the psychological assessment of job hunting and enrollment, it can’t rule out the possibility of answer deceptively under the influence of social desirability. In recent years, with the development of machine learning, new techniques have been developed for personality recognition. This paper uses participants’ self-introduction videos and Big Five personality scale scores to train five-dimensional personality prediction model by key point recording, feature extraction and dimensionality reduction, modeling, and iterative tuning. The results show that the prediction scores of the personality model based on the self-introduction video is close to or achieved medium correlation with scores of questionnaires in all dimensions (r=0.32~0.44, p

Additional file 2: Figure S1. Sample size of tissue samples (A) and plasma sample (B) in Phase I, Phase II and Phase III. The reused samples were showed in dot. Figure S2. Sample sizes of the key categories. Stage I (A) and CA19-9-negative PDAC (B) were shown. Figure S3. Overview of incremental feature selection of PDAC-specific MHB to build PandaX classifiers. Figure S4. Performance of PDAC-CP classifier in differentiating PDAC from CP plasma and covariate analysis. A. ROC of PDACatch classifying PDAC and CP plasma. B. PDACatch scores across different type of samples. In panel B, samples were labelled with cohorts (Train/Validation), pathological types (H: Healthy; CP: chronic pancreatitis) and clinical stages of PDAC (I, II, I-IIA and IIB-IV). For C-F, PDAC-CP classifier’s scores of PDAC and CP plasma samples were grouped by gender (A), age (B), tumour location (C) and tumour size (D). Wilcoxon rank sum test. ns: 0.05 < p

Additional file 1. Supplementary methods.

Building on previous research of Chi and Chi (2022), the current paper revisits estimation in robust structured regression under the $\text{L}_2\text{E}$ criterion. We adopt the majorization-minimization (MM) principle to design a new algorithm for updating the vector of regression coefficients. Our sharp majorization achieves faster convergence than the previous alternating proximal gradient descent algorithm (Chi and Chi, 2022). In addition, we reparameterize the model by substituting precision for scale and estimate precision via a modified Newton's method. This simplifies and accelerates overall estimation. We also introduce distance-to-set penalties to allow constrained estimation under nonconvex constraint sets. This tactic also improves performance in coefficient estimation and structure recovery. Finally, we demonstrate the merits of our improved tactics through a rich set of simulation examples and a real data application., Comment: 30 pages, 8 figures

Spinal root avulsion typically leads to massive motoneuron death and severe functional deficits of the target muscles. Multiple pathological factors such as severe neuron loss, induction of inhibitory molecules, and insufficient regeneration are responsible for the poor functional recovery. Leucine-rich repeat and immunoglobulin-like domain-containing Nogo receptor-interacting protein 1 (LINGO-1), a central nervous system (CNS)-specific transmembrane protein that is selectively expressed on neurons and oligodendrocytes, serves as a potent negative mediator of axonal regeneration and myelination in CNS injuries and diseases. Although accumulating evidence has demonstrated improvement in axonal regeneration and neurological functions by LINGO-1 antagonism in CNS damage, the possible effects of LINGO-1 in spinal root avulsion remain undiscovered. In this study, a LINGO-1 knockdown strategy using lentiviral vectors encoding LINGO-1 short hairpin interfering RNA (shRNA) delivered by the Pluronic F-127 (PF-127) hydrogel was described after brachial plexus avulsion (BPA). We provide evidence that following BPA and immediate reimplantation, transplantation of LINGO-1 shRNA lentiviral vectors encapsulated by PF-127 rescued the injured motoneurons, enhanced axonal outgrowth and myelination, rebuilt motor endplates, facilitated the reinnervation of terminal muscles, improved angiogenesis, and promoted recovery of avulsed forelimbs. Altogether, these data suggest that delivery of LINGO-1 shRNA by a gel scaffold is a potential therapeutic approach for root avulsion. Impact Statement In this study, we attempted transplantation of lentivirus (LV)/leucine-rich repeat and immunoglobulin-like domain-containing Nogo receptor-interacting protein 1 (LINGO-1)-short hairpin interfering RNA (shRNA) encapsulated by the Pluronic F-127 (PF-127) hydrogel into a brachial plexus avulsion (BPA)-reimplantation model. We found that administration of LV/LINGO-1 shRNA facilitates neuron survival and axonal regeneration, attenuates muscle atrophy and motor endplate (MEP) loss, enhances neovascularization, and promotes functional recovery in BPA rats. Co-transplantation of LV/LINGO-1 shRNA and gel reinforces the survival-promoting effect, axonal outgrowth, and angiogenesis in comparison with LV/LINGO-1 shRNA application alone. Our research provides evidence that LV /LINGO-1 shRNA delivered by PF-127 represents a new treatment strategy for BPA repair.

Supplemental material, sj-pdf-1-epb-10.1177_23998083211017909 for Characterizing the complex influence of the urban built environment on the dynamic population distribution of Shenzhen, China, using geographically and temporally weighted regression by Xiaoqian Liu, Bo Huang, Rongrong Li and Jionghua Wang in Environment and Planning B: Urban Analytics and City Science

This paper deals with the grouped variable selection problem. A widely used strategy is to augment the negative log-likelihood function with a sparsity-promoting penalty. Existing methods include the group Lasso, group SCAD, and group MCP. The group Lasso solves a convex optimization problem but is plagued by underestimation bias. The group SCAD and group MCP avoid this estimation bias but require solving a nonconvex optimization problem that may be plagued by suboptimal local optima. In this work, we propose an alternative method based on the generalized minimax concave (GMC) penalty, which is a folded concave penalty that maintains the convexity of the objective function. We develop a new method for grouped variable selection in linear regression, the group GMC, that generalizes the strategy of the original GMC estimator. We present an efficient algorithm for computing the group GMC estimator and also prove properties of the solution path to guide its numerical computation and tuning parameter selection in practice. We establish error bounds for both the group GMC and original GMC estimators. A rich set of simulation studies and a real data application indicate that the proposed group GMC approach outperforms existing methods in several different aspects under a wide array of scenarios., Comment: 34 pages, 4 figures

Picrasma quassioides, a member of the Simaroubaceae family, is the subject of research in numerous pharmacological and chemical studies. This plant mainly contains alkaloids, quassinoids and terpenoids. These molecules exhibit various pharmacological benefits, such as anti-inflammatory, anticancer, and anti-viral effects, on the cardiovascular system. Alkaloids make up the majority of these molecules. This review describes 127 alkaloid substances from P. quassioides. These alkaloids can be divided into the following classes: β-carbolines, canthinones and alkaloid dimers. A compilation of their nuclear magnetic resonance spectroscopy data and possible biosynthetic pathways of these compounds and the pharmacological effects of P. quassioides are also included.

In this paper, novel reduced graphene oxide (rGO) composites (DAPrGOs) modified by diaminopyrene (DAP) were successfully synthesized via a facile solvothermal reaction method and used for supercapacitors. Compared with the pristine rGO, the DAPrGO1 electrode showed distinctly better performance (397.63 F g−1 vs. 80.29 F g−1 of pristine rGO at 0.5 A g−1) with small charge transfer resistance. When a symmetric device was fabricated using DAPrGO1 as the active material, it also exhibited a high capacitance of 82.70 F g−1 at 0.5 A g−1 with an energy density of 25.84 W h kg−1 at a power density of 375 W kg−1, and even offered a high power density of 7500 W kg−1 (18.71 W h kg−1) at 10 A g−1. Moreover, the device possessed good electrochemical stability up to 20 000 cycles, implying promising applications in energy storage fields.

In recent years, deep learning has achieved remarkable fruits in a wide variety of domains, such as recommender systems. In addition, privacy preservation is unprecedentedly necessary in today's era. In this paper, we leverage the privacy preservation problem in recommendation with the deep learning model, i.e., autoencoders. In order to predict user preferences in the collaborative filtering way, the work reconstructs user sparse ratings with autoencoders to estimate unobserved user preferences. To further protect user privacy, the Gaussian mechanism is combined in the stochastic gradient descent process to ensure that the training process meets the requirements of approximate differential privacy.

Facial expression recognition, which plays a vital role in human-machine interaction (HMI), has attracted increasing attention from researchers over the past few years. However, it is still a challenging task for identifying significant discriminative facial features that represent the characteristics of each expression. In this paper, we apply a strategy of curriculum learning (CL) to facial expression recognition to produce effective and discriminative facial representations. Firstly, a learning curriculum is designed by measuring the complexity according to the distance between each sample and the clustering center determined by unsupervised density-distance clustering method in the feature space. Secondly, we develop a multistage training process where a main model is trained by continuously increasing the complexity of training set. The experiment results indicate that our proposed model can achieve a high recognition accuracy of 72.11% on the FER-2013 dataset for seven facial expressions, which outperform other state-of-the-art methods.

Experimental investigation and numerical simulation are performed to obtain the combustion characteristics of premixed hydrogen/air mixtures in a closed container, such as flame propagation velocity, peak pressure and maximum temperature. This paper mainly focused on the influence of initial temperature, hydrogen concentration and steam concentration on hydrogen/air flame propagation. An updated kinetics of hydrogen combustion based on one-step global reaction mechanism was validated by the present experimental data. The numerical results with updated kinetics show a satisfactory agreement with experimental results in term of average flame propagation velocity with the deviation less than 20%. The experimental and simulation results show that the average flame propagation velocity increases with the increase of initial hydrogen concentration and initial temperature, and reduces with the increase of initial steam concentration. The effect of hydrogen concentration on average flame propagation velocity is more significant than that of initial temperature and steam concentration.

For the high-rise building with the basement, it is difficult to identify the position of the embedded fixed end due to the restraints of the backfill soil foundation. The simplified mechanical model of the high-rise building with the basement is established based on several hypotheses, and the mathematical expressions for calculating the elastic lateral displacement and internal forces of the high-rise building with the basement under the inverted-triangle distributed load is then obtained based on the mechanical theory. The influence of parameters and h/H to four variable parameters are researched, and the four variable parameters are the moment M(0)and shear Q(0) at the bottom plate of the basement, the lateral displacement y(h) at the top plate of the basement, the lateral displacement y(H) at the top of the whole structure. Research indicates that the simplified algorithms of the high-rise building with the basement considering the restraint of the backfill soil foundation are reliable, and can be referred and applied by the researchers and engineers.

Ligustri lucidi fructus (LLF) is the fruits of Ligustrum lucidum Ait. (Oleaceae). This review based on nearly 80 literary sources discusses the knowledge of chemistry and biological effects of this species. Several types of chemical constituents considered as the characteristic and active constituents from LLF were isolated including 40 triterpenoids, 48 iridoids, 10 flavones, 10 phenylethanoid glycosides and others. Various extracts and individual compounds derived from this species have been found to possess a variety of pharmacological effects, e.g. anti-tumour, hepatoprotective, immune regulating, antioxidative and anti-ageing effects, anti-inflammation and reducing hypercholesterolaemia effects and so on. The results of data analysis on the chemical, pharmacological characteristics of LLF support the view that this species has many therapeutic properties and indicate its potential as an effective herbal remedy. Finally, some suggestions for further research on chemical and pharmacological properties are given in this review. Theoretical basis was given for further exploiting and utilising LLF.

Rationale Intestinal lymphangiectasia (IL) is a rare enteropathy involving the expansion and rupture of intestinal lymphatic channels. Although several reports have studied cases of primary IL (PIL), this condition is very rare, and is even less commonly encountered in infants. This study aimed to investigate the nutritional therapy and effect assessment of chylous reflux disorder caused by PIL in infants. Patient concerns Infantile patients were enrolled in the Affiliated Beijing Shijitan Hospital of the Capital Medical University between January 2012 and March 2014. The minimum age of onset was 4 months and the maximum age of onset was 16 months, with an average age of 4.9 months. Diagnoses All children were inpatient who had been diagnosed with chylous reflux syndrome (chylothorax and/or chylic abdomen) caused by PIL. Interventions Retrospective analysis and individualized nutrition therapy of these cases were carried out. Finally, nutritional therapy and prognosis of PIL were assessed and summarized. Outcomes All the children survived, showed improvement in the serum total protein, albumin, and HGB levels after nutritional therapy. After comprehensive nutritional therapy, we were able to achieve diarrhea control for all the 9 patients, and after treatment, the children passed soft, yellow stools 1 to 2 times/d. After treatment, the height and weight of all patients increased to within the normal ranges of the World Health Organization standard chart. The mean serum albumin level reached 41.3 g/L. All nutrition-related indicators were found to have significant improvement compared with the baseline levels. Lessons The results revealed that nutritional therapy for the 9 children with PIL was effective, and it may be able to improve the clinical syndromes and symptoms of children with PIL and promote recovery.

Haematopoietic cells arise from spatiotemporally restricted domains in the developing embryo. Although studies of non-mammalian animal and in vitro embryonic stem cell models suggest a close relationship among cardiac, endocardial and haematopoietic lineages, it remains unknown whether the mammalian heart tube serves as a haemogenic organ akin to the dorsal aorta. Here we examine the haemogenic activity of the developing endocardium. Mouse heart explants generate myeloid and erythroid colonies in the absence of circulation. Haemogenic activity arises from a subset of endocardial cells in the outflow cushion and atria earlier than in the aorta-gonad-mesonephros region, and is transient and definitive in nature. Interestingly, key cardiac transcription factors, Nkx2-5 and Isl1, are expressed in and required for the haemogenic population of the endocardium. Together, these data suggest that a subset of endocardial/endothelial cells serve as a de novo source for transient definitive haematopoietic progenitors.

This paper presents the numerical simulation about spray dry purification flue gas of municipal solid wastes (MSW) incineration. The fluent software package, κ-ɛ RNG turbulence model and the stochastic model of the track was chosen to do this work; the research on the flow field characteristics by increasing flow equalization board was carried out. The paper mainly studied the effect for flow characteristics, resistance and coefficient of droplet adhering on the wall. The simulation results are as follows: there was a significant asymmetry of opposite inlet side of the inlet velocity greater than the velocity side without the flow equalization board. The resistance decreases gradually with increasing of the flow equalization board opening diameter. And the coefficient of droplet adhering on the wall firstly increased and then decreased.

Internet video on demand is a new application provided by peer to peer network. Thousands of users are using real multimedia streaming systems like PPlive and PPstream which are the killer applications in the Internet. But jitters are always happening during rush time especially when users are watching less popular videos with low bandwidth network. To improve this, a cache scheduling scheme based on layered coding is proposed in this paper. Our scheme aims at optimizing cache scheduling scheme in peers' local cache while other researches often focus on the proxy server. The available bandwidth considered as energy is allocated by Fermi distribution function for different layer data of FGS, and the prefix buffer is extended dynamically. When the available bandwidth is insufficient, peers are forced to allocate more bandwidth for base layer data, and their prefix buffer is extended so that more base layer data is downloaded for users who watch a video continuously and with interactive operations. Simulation shows that our approach reduces cache missing rate effectively when available bandwidth is insufficient. In local cache, peers replace segments in the cache by formulas of popularity which is proposed to reduce cache missing rate more successful. In general, our scheme can effectively reduce starting delay and times of jitters in a multimedia system with low bandwidth.

As dynamic-level scheduling algorithm does not consider the risk taken by resource nodes and users' demand, it proposes an analytic hierarchy process approach (AHP) based improved DLS algorithm. By such AHP approach, it manages the task provided by users, and obtains the weight value which reflects users' demand. Because the weight value contains the information about deadline, reparation duty and profit of task, it integrates DLS and the weight value to get an improved DLS, called R-DLS which contains the degree of risk taken by resource nodes. Experiments show that R-DLS receives higher average success rate of task execution and profit in one unit of time with task number increasing than DLS.

For the autonomous and selfish features of peers, there are many issues existed in P2P network, such as dynamics of networks, heterogeneity of peers, free-riders, tragedy of commons and malicious behaviors. The topology optimization solutions based on trust are effective methods to solve these problems. However, the trust evaluations in related works only focus on the service trust, or consider both the service trust and the routing trust, but the routing trust evaluation algorithm can not stimulate peers to forward query messages. On this basis, a routing trust evaluation algorithm based on profit sharing is proposed. In the algorithm, according to the contributions of routing recommendation peers, the service demanders share the transaction profit with them in order that the routing recommendation peers can get the routing trust of their neighbors by forwarding query messages. Then, a routing trust aware topology optimization protocol is put forward, which considers both service trust and routing trust. Simulations and analysis show that, the routing trust aware topology optimization protocol can not only promote honest peers to the center of the network, exclude malicious peers and free-riders to the edge of the network, but also improve the query efficiency for honest peers.

The problem of file pollution is prevalent in P2P file sharing system nowadays. The pollution reduces the availability of system dramatically. As the same as immune system, P2P file sharing system is also highly distributive, self-adaptive and self-organized. Artificial immunology based anti-pollution object reputation mechanism is established to select neighbor peer set and improve the system availability. Vector space similarity is computed to weight the votes collected from other peers. Horting graph-theoretic approach is employed to deal with the data sparsity problem. Self-adaptive reputation threshold scheme is adopted to estimate the file authenticity. Simulation results show that the system can distinguish between good and polluted files accurately and thwart the proliferation of pollution in the network with low communication cost.