Search

IntroductionPneumonia-induced sepsis can cause multiple organ dysfunction including acute lung and kidney injury (ALI and AKI). Surfactant protein A (SP-A), a critical innate immune molecule, is expressed in the lung and kidney. Extracellular vesicles like exosomes are involved in the processes of pathophysiology. Here we tested one hypothesis that SP-A regulates pneumonia-induced AKI through the modulation of exosomes and cell death.MethodsWild-type (WT), SP-A knockout (KO), and humanized SP-A transgenic (hTG, lung-specific SP-A expression) mice were used in this study.ResultsAfter intratracheal infection with Pseudomonas aeruginosa, KO mice showed increased mortality, higher injury scores, more severe inflammation in the lung and kidney, and increased serum TNF-α, IL-1β, and IL-6 levels compared to WT and hTG mice. Infected hTG mice exhibited similar lung injury but more severe kidney injury than infected WT mice. Increased renal tubular apoptosis and pyroptosis in the kidney of KO mice were found when compared with WT and hTG mice. We found that serum exosomes from septic mice cause ALI and AKI through mediating apoptosis and proptosis when mice were injected intravenously. Furthermore, primary proximal tubular epithelial cells isolated from KO mice showed more sensitivity than those from WT mice after exposure to septic serum exosomes.DiscussionCollectively, SP-A attenuates pneumonia-induced ALI and AKI by regulating inflammation, apoptosis and pyroptosis; serum exosomes are important mediators in the pathogenesis of AKI.

Osteoarthritis (OA), viewing as a degenerative aseptic inflammatory disease, is characterized by joint pain and inflammation that significantly affects the quality of patients’ life, especially for the elder. Although rapid progress has been achieved in elucidating the underlying mechanisms of OA occurrence and progression, there is still a lack of effective clinical therapeutics for OA patients. Currently the most common treatments including drug therapy and surgical operations are not very satisfactory in majority of cases, so it is worthy to explore new remedies. During the past few decades, a number of novel forms of regulated cell death have been reported widely, typified by ferroptosis, with its prominent features including reactive oxygen species (ROS) elevation, lipid peroxidation, iron accumulation and glutathione deprivation. Our study was designed to identify the functional roles of differentially expressed ferroptosis-related genes in OA, which were screened out by referring to GEO database via bioinformatics analyses. Human chondrocytes were applied to validate the above findings in the scenario of ferroptosis inhibitors administration. Results partially proved the consistency with bioinformatics analyses that ATF3 and TFRC were highly expressed in interleukin-1β (IL-1β)-stimulated chondrocytes whereas CXCL2 and JUN were downregulated. Besides, TFRC was firstly validated to be upregulated in IL-1β-stimulated chondrocytes, which could be reversed by ferroptosis inhibitors. In conclusion, our study reported two prominent ferroptosis-related genes, ATF3 and TFRC are upregulated in IL-1β-stimulated chondrocytes while CXCL2 and JUN are downregulated. And preliminary results demonstrated that TFRC might serve as an accomplice of ferroptosis process in IL-1β-stimulated chondrocytes and ferroptosis inhibitors have the potential to inhibit ROS in IL-1β-stimulated chondrocytes.

Additional file 1. Table S1

Machine learning force fields (MLFF) have been proposed to accelerate molecular dynamics (MD) simulation, which finds widespread applications in chemistry and biomedical research. Even for the most data-efficient MLFFs, reaching chemical accuracy can require hundreds of frames of force and energy labels generated by expensive quantum mechanical algorithms, which may scale as $O(n^3)$ to $O(n^7)$, with $n$ proportional to the number of basis functions. To address this issue, we propose a multi-stage computational framework -- ASTEROID, which lowers the data cost of MLFFs by leveraging a combination of cheap inaccurate data and expensive accurate data. The motivation behind ASTEROID is that inaccurate data, though incurring large bias, can help capture the sophisticated structures of the underlying force field. Therefore, we first train a MLFF model on a large amount of inaccurate training data, employing a bias-aware loss function to prevent the model from overfitting tahe potential bias of this data. We then fine-tune the obtained model using a small amount of accurate training data, which preserves the knowledge learned from the inaccurate training data while significantly improving the model's accuracy. Moreover, we propose a variant of ASTEROID based on score matching for the setting where the inaccurate training data are unlabeled. Extensive experiments on MD datasets and downstream tasks validate the efficacy of ASTEROID. Our code and data are available at https://github.com/abukharin3/asteroid.

Additional file 9. Supplementary fig 1

As a crucial building block in vertical Federated Learning (vFL), Split Learning (SL) has demonstrated its practice in the two-party model training collaboration, where one party holds the features of data samples and another party holds the corresponding labels. Such method is claimed to be private considering the shared information is only the embedding vectors and gradients instead of private raw data and labels. However, some recent works have shown that the private labels could be leaked by the gradients. These existing attack only works under the classification setting where the private labels are discrete. In this work, we step further to study the leakage in the scenario of the regression model, where the private labels are continuous numbers (instead of discrete labels in classification). This makes previous attacks harder to infer the continuous labels due to the unbounded output range. To address the limitation, we propose a novel learning-based attack that integrates gradient information and extra learning regularization objectives in aspects of model training properties, which can infer the labels under regression settings effectively. The comprehensive experiments on various datasets and models have demonstrated the effectiveness of our proposed attack. We hope our work can pave the way for future analyses that make the vFL framework more secure., Comment: 9 pages

Additional file 2. Table S2

Additional file 3. Table S3

At present, the traffic problem is a problem that the government attaches great importance to. Many papers also put forward their own visualization models for traffic problems. This research focused on the Map-matching and Spatial-temporal Visualization of Expressway Traffic Accident Information and improves the original two-dimensional visual model of accident rate into a three-dimensional model. The goal is to represent more attributes in a visual model and make them easier to compare, so as to provide users with more intuitive visual information.

Biomarker development for clinical checkpoint inhibition is still in its early stages. It is critical to determine the cause of the lack of a long-term response in patients after immune checkpoint blockade (ICB) treatment and to develop composite biomarkers or signatures to improve personalized approaches. Three modules that were significantly correlated with the immunotherapeutic response were identified. Stimulatory pathways of cellular immunity, extracellular matrix formation-related pathways, and ATP metabolism-related pathways were enriched. Two distinct transcriptional subtypes were determined. Tumor microenvironment (TME) characteristics were highly correlated with “hot” and “cold” tumors. The ICB score was significantly correlated with clinical characteristics including age, Breslow depth, Clerk level, AJCC stage, and T stage. Meanwhile, a low ICB score is characterized by increased activation of immunity, a higher level of immune infiltration, and immune molecule expression. The ICB score showed a robust ability to predict melanoma prognosis in the discovery, internal validation, and external validation cohorts. In addition, a low ICB score was linked to a higher CR/PR rate in the immunotherapeutic cohort. The ICB score could reflect the pre-existing immune features and the expression pattern of “Cold” versus “Hot” tumors in melanoma patients. Thus, it has the potential to serve as a reliable predictor of melanoma prognosis and response to ICB therapy.

Additional file 2.

Thanks to the capacity for long-range dependencies and robustness to irregular shapes, vision transformers and deformable convolutions are emerging as powerful vision techniques of segmentation.Meanwhile, Graph Convolution Networks (GCN) optimize local features based on global topological relationship modeling. Particularly, they have been proved to be effective in addressing issues in medical imaging segmentation tasks including multi-domain generalization for low-quality images. In this paper, we present a novel, effective, and robust framework for medical image segmentation, namely, UGformer. It unifies novel transformer blocks, GCN bridges, and convolution decoders originating from U-Net to predict left atriums (LAs) and LA scars. We have identified two appealing findings of the proposed UGformer: 1). an enhanced transformer module with deformable convolutions to improve the blending of the transformer information with convolutional information and help predict irregular LAs and scar shapes. 2). Using a bridge incorporating GCN to further overcome the difficulty of capturing condition inconsistency across different Magnetic Resonance Images scanners with various inconsistent domain information. The proposed UGformer model exhibits outstanding ability to segment the left atrium and scar on the LAScarQS 2022 dataset, outperforming several recent state-of-the-arts.

Point process models are of great importance in real world applications. In certain critical applications, estimation of point process models involves large amounts of sensitive personal data from users. Privacy concerns naturally arise which have not been addressed in the existing literature. To bridge this glaring gap, we propose the first general differentially private estimation procedure for point process models. Specifically, we take the Hawkes process as an example, and introduce a rigorous definition of differential privacy for event stream data based on a discretized representation of the Hawkes process. We then propose two differentially private optimization algorithms, which can efficiently estimate Hawkes process models with the desired privacy and utility guarantees under two different settings. Experiments are provided to back up our theoretical analysis.

We investigate the role of noise in optimization algorithms for learning over-parameterized models. Specifically, we consider the recovery of a rank one matrix $Y^*\in R^{d\times d}$ from a noisy observation $Y$ using an over-parameterization model. We parameterize the rank one matrix $Y^*$ by $XX^\top$, where $X\in R^{d\times d}$. We then show that under mild conditions, the estimator, obtained by the randomly perturbed gradient descent algorithm using the square loss function, attains a mean square error of $O(\sigma^2/d)$, where $\sigma^2$ is the variance of the observational noise. In contrast, the estimator obtained by gradient descent without random perturbation only attains a mean square error of $O(\sigma^2)$. Our result partially justifies the implicit regularization effect of noise when learning over-parameterized models, and provides new understanding of training over-parameterized neural networks.

We have previously reported that preconception exposure to iAs may contribute to the development of diabetes in mouse offspring by altering gene expressions in paternal sperm. However, the individual contributions of iAs and its methylated metabolites, monomethylated arsenic (MAs) and dimethylated arsenic (DMAs), to changes in the sperm transcriptome could not be determined because all three As species are present in sperm after in vivo iAs exposure. The goal of the present study was to assess As species-specific effects using an ex vivo model. We exposed freshly isolated mouse sperm to either 0.1 or 1 μM arsenite (iAs(III)) or the methylated trivalent arsenicals, MAs(III) and DMAs(III), and used RNA-sequencing to identify differentially expressed genes, enriched pathways, and associated protein networks. For all arsenicals tested, the exposures to 0.1 μM concentrations had greater effects on gene expression than 1 μM exposures. Transcription factor AP-1 and B cell receptor complexes were the most significantly enriched pathways in sperm exposed to 0.1 μM iAs(III). The Mre11 complex and Antigen processing were top pathways targeted by exposure to 0.1 μM MAs(III) and DMAs(III), respectively. While there was no overlap between gene transcripts altered by ex vivo exposures in the present study and those altered by in vivo exposure in our prior work, several pathways were shared, including PI3K-Akt signaling, Focal adhesion, and Extracellular matrix receptor interaction pathways. Notably, the protein networks associated with these pathways included those with known roles in diabetes. This study is the first to assess the As species-specific effects on sperm transcriptome, linking these effects to the diabetogenic effects of iAs exposure.

The role of Ni-Ag bimetallic magnetic catalysts in the catalytic transfer of hydrogenated soybean oil was studied. First, a Ni-Ag0.15/PVP-DB-171/SiO2/Fe3O4 magnetic catalyst with a magnetic saturation value of 10.431 emu / g was prepared. It was found that the addition of the metal Ag promoter enhanced the dispersion of Ni on the PVP-DB-171/SiO2/Fe3O4 support. The conditions of the catalytic transfer hydrogenation (CTH) (temperature 80°C, catalyst loading 0.23%, donor concentration 0.32 mol /50 mL H2O, and time 90 min) showed the effects of the bimetallic catalysts on the soybean oil hydrogenation process. The hydrogenated soybean oil linolenic acid, linoleic acid and oleic acid reaction rate constants were 4.95×10-2, 8.6×10-3 and 7.54×10-4, respectively. The selectivity of linolenic acid and linoleic acid is as high as 5.75 and 11.4, respectively; the iodine value (IV) of soybean oil after hydrogenation is 102 g I2/100g and the trans fatty acids(TFAs) content is only 1.7%. The use efficiency of the catalyst decreased to 60% after 8 cycles. Catalytic transfer hydrogenation has important research significance and application prospects for the preparation of low-trans hydrogenated oils and fats. This method also provides a theoretical basis for the development of the oil hydrogenation industry.

This paper addresses sparse signal reconstruction under various types of structural side constraints with applications in multi-antenna systems. Side constraints may result from prior information on the measurement system and the sparse signal structure. They may involve the structure of the sensing matrix, the structure of the non-zero support values, the temporal structure of the sparse representationvector, and the nonlinear measurement structure. First, we demonstrate how a priori information in form of structural side constraints influence recovery guarantees (null space properties) using L1-minimization. Furthermore, for constant modulus signals, signals with row-, block- and rank-sparsity, as well as non-circular signals, we illustrate how structural prior information can be used to devise efficient algorithms with improved recovery performance and reduced computational complexity. Finally, we address the measurement system design for linear and nonlinear measurements of sparse signals. Moreover, we discuss the linear mixing matrix design based on coherence minimization. Then we extend our focus to nonlinear measurement systems where we design parallel optimization algorithms to efficiently compute stationary points in the sparse phase retrieval problem with and without dictionary learning.

Deep learning based camera pose estimation from monocular camera images has seen a recent uptake in Visual SLAM research. Even though such pose estimation approaches have excellent results in small confined areas like offices and apartment buildings, they tend to do poorly when applied to larger areas like outdoor settings, mainly because of the scarcity of distinctive features. We propose GODSAC* as a camera pose estimation approach that augments pose predictions from a trained neural network with noisy odometry data through the optimization of a pose graph. GODSAC* outperforms the state-of-the-art approaches in pose estimation accuracy, as we demonstrate in our experiments.

In this thesis, we study the provenance of querying graph databases. Compared to using the semiring of polynomials as the most general form of provenance for relational databases (Green, Karvounarakis, & Tannen, 2007), we show that the most general provenance for querying graph databases can be represented by regular expressions over paths in the database. In this work we focus on the single-source provenance, which is a more general representation and contains more information than the single-source, single-target problem considered in (Ramusat, Maniu, & Senellart, 2018). We present an algorithm that computes single-source multidimensional provenances for graph databases, where each dimension represents an application provenance semiring, and also propose a potential application, by using parse tree techniques and deriving results for various application provenances.

Improved understanding of the genetic structure and genetic variation patterns in natural populations is an essential element of breeding programs that focus on forest species. Castanopsis fissa is an important, fast-growing evergreen species in southern China, where it is used for reforestation. We sampled 28 different C. fissa populations from three provinces in southern China and investigated their genetic variation and genetic relationship through an analysis of sequence-related amplified polymorphism (SRAP) markers. We acquired 236 electrophoretic bands from the 241 individuals analyzed by SRAP; 195 bands were polymorphic. Analysis of molecular variance (AMOVA) detected significant genetic variation within and among the populations; genetic variation among populations was smaller than variation within populations (p

Generative Adversarial Imitation Learning (GAIL) is a powerful and practical approach for learning sequential decision-making policies. Different from Reinforcement Learning (RL), GAIL takes advantage of demonstration data by experts (e.g., human), and learns both the policy and reward function of the unknown environment. Despite the significant empirical progresses, the theory behind GAIL is still largely unknown. The major difficulty comes from the underlying temporal dependency of the demonstration data and the minimax computational formulation of GAIL without convex-concave structure. To bridge such a gap between theory and practice, this paper investigates the theoretical properties of GAIL. Specifically, we show: (1) For GAIL with general reward parameterization, the generalization can be guaranteed as long as the class of the reward functions is properly controlled; (2) For GAIL, where the reward is parameterized as a reproducing kernel function, GAIL can be efficiently solved by stochastic first order optimization algorithms, which attain sublinear convergence to a stationary solution. To the best of our knowledge, these are the first results on statistical and computational guarantees of imitation learning with reward/policy function approximation. Numerical experiments are provided to support our analysis.

Ni–Agx/PVP-DB-171/SiO2/Fe3O4 magnetic nanoparticle (NP) catalysts were prepared using PVP-DB-171/SiO2/Fe3O4 magnetic NPs as a support. The prepared catalysts were characterized via X-ray powder diffraction, scanning electron microscopy/energy dispersive X-ray spectroscopy, vibrating sample magnetometer, inductively coupled plasma-optical emission spectroscopy and N2 adsorption–desorption isotherm. The results indicate that the dispersion of nickel on the PVP-DB-171/SiO2/Fe3O4NP support was improved by incorporating Ag as a promoter. The optimal Ni–Ag molar ratio was 0.15. Soybean oil hydrogenation was carried out with Raney nickel, Ni/PVP-DB-171/SiO2/Fe3O4 and Ni–Ag0.15/PVP-DB-171/SiO2/Fe3O4 catalysts. The results show that the Ni–Ag0.15/PVP-DB-171/SiO2/Fe3O4 magnetic NP catalyst has the highest activity and stability. The use efficiency of the catalyst decreased to 60% after 8 cycles. The content of trans fatty acids in hydrogenated soybean oil decreased to 10.4% after 90 min of hydrogenation. Compared with filtration or centrifugation, the magnetic NP catalyst can be quickly and cost-efficiently separated by applying an external magnetic field.

Candida antarctica lipase B (CALB) was immobilized on Fe3O4/SiOx-g-P(GMA) polymer carrier to catalyzed the transesterification of soybean oil and phytosterol. The enzyme loading of the obtained particles was 98.7 mg/g supports and the enzyme activity was 1226.5 U/g. The average particle size was 100.5 ± 1.30 nm and the magnetization was 15.80 emu/g. The immobilized enzyme showed higher activities at a wider range of pH and temperatures. Its optimum reaction temperature was up to 50 °C; increased by 5 °C compared to the free enzyme. The obtained magnetic immobilized Fe3O4/SiOx-g-P(GMA) lipase was nanoscale. First-grade soybean oils were used as a substrate. System pH was adjusted to 7.0. The optimal reaction temperature was 50 °C and the reaction time was 3 h. The phytosterol concentration of 5% and immobilized CALB of 2% were obtained. The conversion rate of transesterification reaction between soybean oil and phytosterol was 86.2%. The use of magnets can quickly separate the immobilized enzymes from the substrates. The relative activity of the immobilized enzymes was 83.0% when reused seven times. The prepared immobilized CALB can improve efficiently enzyme activity and reutilization.

Anti-ram bollard systems, which are installed around buildings and infrastructure, can prevent unauthorized vehicles from entering, maintain distance from vehicle-borne improvised explosive devices (VBIED) and reduce the corresponding damage. Compared with a fixed bollard system, a removable bollard system provides more flexibility as it can be removed when needed. This paper first proposes a new type of K4-rated removable anti-ram bollard system. To simulate the collision of a vehicle hitting the bollard system, a finite element model was then built and verified through comparison of numerical simulation results and existing experimental results. Based on the orthogonal design method, the factors influencing the safety and economy of this proposed system were examined and sorted according to their importance. An optimal design scheme was then produced. Finally, to validate the effectiveness of the proposed design scheme, four dynamic impact tests, including two front impact tests and two side impact tests, have been conducted according to BSI Specifications. The residual rotation angles of the specimen are smaller than 30o and satisfy the requirements of the BSI Specification.

Numerous empirical evidence has corroborated that the noise plays a crucial rule in effective and efficient training of neural networks. The theory behind, however, is still largely unknown. This paper studies this fundamental problem through training a simple two-layer convolutional neural network model. Although training such a network requires solving a nonconvex optimization problem with a spurious local optimum and a global optimum, we prove that perturbed gradient descent and perturbed mini-batch stochastic gradient algorithms in conjunction with noise annealing is guaranteed to converge to a global optimum in polynomial time with arbitrary initialization. This implies that the noise enables the algorithm to efficiently escape from the spurious local optimum. Numerical experiments are provided to support our theory., Comment: International Conference on Machine Learning (ICML), 2019

Stochastic simulation has been widely used to analyze the performance of complex stochastic systems and facilitate decision making in those systems. Stochastic simulation is driven by the input model, which is a collection of probability distributions that model the stochasticity in the system. The input model is usually estimated using a finite amount of data, which introduces the so-called input model uncertainty to the simulation output. How to quantify input uncertainty has been studied extensively, and many methods have been proposed for the batch data setting, i.e., when all the data are available at once. However, methods for "streaming data" arriving sequentially in time are still in demand, despite that streaming data have become increasingly prevalent in modern applications. To fill this gap, we propose a two-layer importance sampling framework that incorporates streaming data for online input uncertainty quantification. Under this framework, we develop two algorithms that suit different application scenarios: the first scenario is when data come at a fast speed and there is no time for any new simulation in between updates; the second is when data come at a moderate speed and a few but limited simulations are allowed at each time stage. We prove the consistency and asymptotic convergence rate results, which theoretically show the efficiency of our proposed approach. We further demonstrate the proposed algorithms on a numerical example of the news vendor problem.

Asynchronous momentum stochastic gradient descent algorithms (Async-MSGD) is one of the most popular algorithms in distributed machine learning. However, its convergence properties for these complicated nonconvex problems is still largely unknown, because of the current technical limit. Therefore, in this paper, we propose to analyze the algorithm through a simpler but nontrivial nonconvex problem - streaming PCA, which helps us to understand Aync-MSGD better even for more general problems. Specifically, we establish the asymptotic rate of convergence of Async-MSGD for streaming PCA by diffusion approximation. Our results indicate a fundamental tradeoff between asynchrony and momentum: To ensure convergence and acceleration through asynchrony, we have to reduce the momentum (compared with Sync-MSGD). To the best of our knowledge, this is the first theoretical attempt on understanding Async-MSGD for distributed nonconvex stochastic optimization. Numerical experiments on both streaming PCA and training deep neural networks are provided to support our findings for Async-MSGD., Comment: arXiv admin note: text overlap with arXiv:1802.05155

Momentum Stochastic Gradient Descent (MSGD) algorithm has been widely applied to many nonconvex optimization problems in machine learning, e.g., training deep neural networks, variational Bayesian inference, and etc. Despite its empirical success, there is still a lack of theoretical understanding of convergence properties of MSGD. To fill this gap, we propose to analyze the algorithmic behavior of MSGD by diffusion approximations for nonconvex optimization problems with strict saddle points and isolated local optima. Our study shows that the momentum helps escape from saddle points, but hurts the convergence within the neighborhood of optima (if without the step size annealing or momentum annealing). Our theoretical discovery partially corroborates the empirical success of MSGD in training deep neural networks., Comment: arXiv admin note: text overlap with arXiv:1806.01660

// Ning Ma 1 , Yi Wu 1 , Fei Xie 1 , Kexin Du 1 , Yuan Wang 1 , Linxin Shi 1 , Linbao Ji 1 , Tianyi Liu 1 and Xi Ma 1, 2 1 State Key Laboratory of Animal Nutrition, China Agricultural University, Beijing 100193, China 2 Departments of Internal Medicine and Biochemistry, University of Texas Southwestern Medical Center, Dallas, TX 75390-9113, USA Correspondence to: Xi Ma, email: maxi@cau.edu.cn Keywords: dimethyl fumarate, ultraviolet radiation, microbial composition and distribution, intestinal barrier function, mycotoxin Received: February 09, 2017 Accepted: April 27, 2017 Published: May 16, 2017 ABSTRACT The effects of dimethyl fumarate (DMF) on mycotoxins and animal growth performance are well documented. However, its mechanism of anti-mildew effects is still unknown. The current study investigated how DMF detoxified the mycotoxin and improved the growth performance using BALB/c mice model, especially its effects on intestinal barrier function and gut micro-ecology. Our study also compared with the ultraviolet radiation (UR) treatment, a traditional anti-mildew control (TC). The results indicated that the DMF treatment had a lower contents of mycotoxin, better growth performance and improved mucosal morphology ( P < 0.05), accompanied with the decreased intestinal permeability and the tighter gut barrier. Moreover, the efficiency of DMF was better than TC ( P < 0.05). 16S rRNA gene sequence analysis revealed that the richness and diversity of bacteria was increased in DMF treatment. The most abundant OTUs belonged to Firmicutes and Bacteroidetes, and their changes in DMF were more moderate than the TC group, suggesting a more stable micro-ecology and the positive impact of DMF on the biodiversity of intestine. Specifically, the increased abundance of bacteria producing short-chain fatty acids (SCFAs), such as Gemella, Roseburia, Bacillus and Bacteroides in DMF group and prebiotics such as Lactobacillus in TC group, suggested a more healthier microbial composition and distribution. These findings supported that DMF had significant effects on animal’s growth performance and intestinal barrier function by modulating the pathway of nutrient absorption and increasing the diversity and balance of gut microbes, which also illuminate that DMF is more efficient than traditional anti-mildew method.

Loblolly pine (Pinus taeda L.) is one of the most important species for oleoresin (a mixture of terpenoids) in South China. The high oleoresin content of loblolly pine is associated with resistance to bark beetles and other economic benefits. In this study, we conducted transcriptome analyses of loblolly pine secondary xylem to gain insight into the genes involved in terpenoid biosynthesis. A total of 372 unigenes were identified as being critical for oleoresin production, including genes for ATP-binding cassette (ABC) transporters, the cytochrome P450 (CYP) protein family, and terpenoid backbone biosynthesis enzymes. Six key genes involved in terpenoid biosynthetic pathways were selected for multiple sequence alignment, conserved motif prediction, and phylogenetic and expression profile analyses. The protein sequences of all six genes exhibited a higher degree of sequence conservation, and upstream genes were relatively more conserved than downstream genes in terpenoid biosynthetic pathways. The N-terminal regions of these sequences were less conserved than the C-terminal ends, as the N-terminals were quite diverse in both length and composition. The phylogenetic analyses revealed that most genes originated from gene duplication after species divergence, and partial genes exhibited incomplete lineage sorting. In addition, the expression profile analyses showed that all six genes exhibited high expression levels during the high-oleoresin-yielding phase.

Genetic parameters for height (H), diameter at breast height (DBH), stem straightness (STR), and under crown clear bole height (CH) of loblolly pine (Pinus taeda L.) were estimated for 255 families (209 open pollinated (OP) and 46 controlled pollinated (CP) families) using a family model and an individual tree model at age 1, 2, 3, 5, 11, and 15 years. Heritability estimates for growth traits of individual trees at age 11 years were the highest (0.17-0.78), and those at age 15 years were the lowest (0.05-0.74). Heritability estimates for DBH, STR, and CH were lower than those for H. Genetic correlations between H and DBH were generally strongly positive, attained a maximum values at age 2 to 3, and declined slightly thereafter. The genetic correlations between CH at age 11 and both H and DBH at different ages were moderate. Age-age genetic correlations for growth traits were moderate to high (0.56-0.91) at age 5 for half-rotation age (15 years), indicating the opportunity exists for early selection. Indirect selection from the age 5 to 11 years for H and DBH could be expected to produce gains of over 50% and 35% respectively, for these two ages, relative to direct selection at age 15. Efficiencies of early selection for H and DBH indicated that growth at maturity could be improved by early selection.

Quantitative trait locus (QTL) detection was carried out for growth traits in 122 F1 progenies of Pinus elliottii var. elliottii (PEE) × P. caribaea var. hondurensis (PCH) hybrid. The PCH male parent linkage map contained 108 markers in 16 linkage groups, while the PEE female parent contained 93 markers in 19 linkage groups. Sequence-related amplified polymorphism (SRAP), microsatellite (SSR), expressed sequence tag polymorphism (ESTP) and inter-simple sequence repeat (ISSR) were selected from an existing linkage map. Growth traits investigated were height at age five (HT05) and six years (HT06), diameter at breast height at age five (DBH05) and six years (DBH06) and annual growth from age five to six years of height (AGHT) and diameter at breast height (AGDBH). Kruskal-Wallis and interval mapping approaches were used to estimate levels of significance, the number of QTLs, the percentage of the phenotypic variation explained by each of QTLs and their positions on the genetic linkage maps. Twenty six QTLs with significance levels p < 0.05 were detected on the parental maps for the six growth traits, which explained more than 15% of the phenotypic variation, suggesting an existence of major-effect genes. Several QTLs had the pleiotropy controlling more than one single growth trait. Overall, the proportion of phenotypic variation explained by QTLs ranged from 5.9% to 40.6% for HT05 and HT06, from 6.6% to 42.0% for DBH05 and DBH06 and from 5.8% to 22.3% for AGHT and AGDBH. The results from this study provide a basis for marker-aided selection (MAS) in PEE × PCH hybrids.

A total of 122 F1 individuals from a single full-sib interspecific hybrid family crossed between Pinus elliottii var. elliottii (PEE) and P. caribaea var. hondurensis (PCH) were used to construct a detailed genetic linkage maps using four types of molecular markers: sequence-related amplified polymorphism (SRAP), microsatellite (SSR), expressed sequence tag polymorphism (ESTP) and inter-simple sequence repeat (ISSR). There were 381 SRAP, 108 SSR, 25 ESTP and 32 ISSR loci, segregating in the interspecific F1 hybrid individuals. Framework maps were constructed at a LOD score threshold of 4.0 using the JoinMap® 3.0. The map for the male parent (PCH) had 108 markers in 16 linkage groups (LGs), with a total length of 1,065.9 cM (Kosambi) and an average marker interval of 9.87 cM. The map for the female parent (PEE) contained 93 markers in 19 LGs, with a total length of 1,006.7 cM (Kosambi) and an average marker interval of 10.82 cM. The maps for PCH and PEE covered 56.5 and 70.3 % of their respective genomes. Based on the position of 36 loci segregating in both parents, 8 homologous LGs between PEE and PCH were identified.

Light Chaser Animation is an animation studio based in Beijing, and our goal is to create world-class CG animation content. Sent is our second work in virtual reality and our first VR story short film.Lately, some key figures in the CG field have questioned the viability of narrative storytelling in the VR medium. Sent is our first attempt in answering the question.This 5-minute, 40-second short tells the story of an emoji "Goodbye." The narrative begins inside the phone of a boy who is having a quarrel with his girlfriend. Behind the chat screen, all kinds of emoji characters are gathered and vie to be chosen and applied to the conversation. When the boy selects "Goodbye," the emoji embarks upon a magical journey that would turn things in surprising ways for all.

Wound healing is an intricate process whereby the skin, or similar tissues, repairs itself after injury. Impaired wound healing with loss of temporospatial orchestration of processes remains challenging and can be highly debilitating to patients. Recently, the novel strategy of administering mesenchymal stem cells (MSCs) derived from bone marrow, adipose tissues and other sources, has achieved high therapeutic efficiency in healing wounds. Adipose-derived stem cells (ASCs) are pluripotent stem cells with the ability to differentiate into different lineages and to secrete paracrine factors to initiate tissue regeneration. However, the exact molecular mechanism by which ASCs exert their function is not well characterized. MicroRNAs (MiRNAs) are important regulators of diverse cell functions and the cytokine network is necessary for endothelial cell migration, capillary formation, matrix metalloproteinase production, collagen synthesis and degradation. This review will focus on the significant role of MicroRNA within ASCs and how these are interlinked to the pathogenesis of wound healing.

Anterior cervical hypertrophic scars caused by severe burn are prone to contracture deformation. Even after multiple skin graft procedures, limitation of neck motion still occurs, especially in patients with hypertrophic scarring. This study examines the feasibility of associating the free scapular flap and platysmaplasty for reconstruction of recurrent neck contracture. Patients with severe scar contracture after multiple skin grafting and with hypertrophic scarring were under investigation. After complete release of the anterior cervical scar, a transection of platysma combined with suture fixation of platysma muscle flap to the surface of chin bone was performed, and the vascular anastomotic free scapular flap was covered. Functional exercise was strengthened postoperatively. All flaps (12 cases) survived well with obvious improvement of neck motion and satisfactory appearance. Free scapular flap associated to platysmaplasty is one of the preferred alternatives for scar reconstruction in patients with recurrent neck contracture or severe hypertrophic scarring. Level of Evidence: Level IV, therapeutic study.

OBJECTIVE Some patients experience massive neglected scalp malignant tumors involving the soft tissues and bone. Management of these tumors includes thorough tissue resection to obtain negative margins and cover the defects with enough composite tissues. In this article, we presented our experiences of reconstructions with free anterolateral thigh (ALT) flaps. METHODS We performed a retrospective review of 6 patients who had a single-stage tumor removal and free-tissue transfer over the last 10 years. All the patients were treated with ALT flaps with or without fascia lata according to the defects. RESULTS The size of the flaps ranged from 8 to 25 cm in length and from 6 to 20 cm in width. All patients healed well without major complications except for 1 case that partial skin-graft loss occurred at the donor sites. CONCLUSIONS Scalp malignant tumors are generally associated with high mortality rates, and the reconstruction surgery is quite complicated. We demonstrate that free ALT flap transfer yields acceptable survival with favorable safety.

To establish a technique for moderate and severe blepharoptosis to exert the function of residual levator muscle, 30 patients (38 eyes) who had moderate and severe blepharoptosis were treated, and the results including complications were followed up and valued. Operation was performed via anterior transcutaneous incision. After separating levator muscle and tarsus, the amount excision of levator muscle and tarsus was accurately estimated. Part of levator muscle associated with tarsus was excised. The position and contour of the upper lid margin, uncovering of the pupils, and symmetry of the palpebral fissures were assessed. Incidence of postoperative complications, including infection, was also considered. Among the 38 eyes, all had a good or fair outcome except for 2 eyes. Eyelid was raised to normal level by this technique with a natural eyelid contour. Some cases presented slight neuropathies in the early stage after operation. We concluded that excision of levator muscle and tarsus could be applied to correct moderate and severe blepharoptosis with some residual levator function, which could achieve good aesthetic results ultimately.

Quantitative trait locus (QTL) analysis was conducted on three full-sib Pinus elliottii var. elliottii × P. caribaea var. hondurensis (PEE × PCH) hybrid families to identify genomic regions affecting two growth traits, height (Ht) and DBH at age 5, and seven morphological (“form”) traits. The QTL detection population consisted of family A containing 122 progeny, which was related to the validation population family B (n = 66) via a common PEE parent, but unrelated to the second validation population, family C (n = 107). Using a double pseudo-testcross QTL mapping approach, quantitative trait loci (QTLs) had been identified in the detection population for Ht, DBH, and three form traits. One QTL for Ht05 was validated in the shared common PEE lineage. For two other form traits, QTLs were validated by identifying QTLs that co-located to the same genomic regions in both validation populations. Two QTL, for another form trait and DBH05 were not identified in the original detection population but occurred in both validation populations. The QTL positions detected in this study provide a resource to identify form-trait loci using the Pinus genome or expressed sequence tag (EST) sequences, which may be useful for future association studies.

Deep learning refers to the shining branch of machine learning that is based on learning levels of representations. Convolutional Neural Networks (CNN) is one kind of deep neural network. It can study concurrently. In this article, we gave a detailed analysis of the process of CNN algorithm both the forward process and back propagation. Then we applied the particular convolutional neural network to implement the typical face recognition problem by java. Then, a parallel strategy was proposed in section4. In addition, by measuring the actual time of forward and backward computing, we analysed the maximal speed up and parallel efficiency theoretically., Comment: 10 pages, 6 figures

The complete chloroplast sequence of Gnetum montanum (GenBank accession number: NC_021438.1) was determined and characterized in this study. The size of the chloroplast genome of Gnetum montanum is 115 019 bp. Including a large single copy (LSC) region of 66 697 bp and a small single copy (SSC) region of 9494 bp separated by a pair of identical inverted repeat regions (IRs) of 19 414 bp each. A total of 108 genes that were successfully annotated contains 65 protein-coding genes, 40 tRNA genes and 3 rRNA genes. The GC content of the chloroplast genome of Gnetum montanum is 38.16%. Twelve genes contain one introns.