Deep Reinforcement Learning in Nonstationary Environments With Unknown Change Points.

Deep reinforcement learning (DRL) is a powerful tool for learning from interactions within a stationary environment where state transition and reward distributions remain constant throughout the process. Addressing the practical but challenging nonstationary environments with time-varying state transition or reward function changes during the interactions, ingenious solutions are essential for the stability and robustness of DRL agents. A key assumption to cope with nonstationary environments is that the change points between the previous and the new environments are known beforehand. Unfortunately, this assumption is impractical in many cases, such as outdoor robots and online recommendations. To address this problem, this article presents a robust DRL algorithm for nonstationary environments with unknown change points. The algorithm actively detects change points by monitoring the joint distribution of states and actions. A detection boosted, gradient-constrained optimization method then adapts the training of the current policy with the supporting knowledge of formerly well-trained policies. The previous policies and experience help the current policy adapt rapidly to environmental changes. Experiments show that the proposed method accumulates the highest reward among several alternatives and is the fastest to adapt to new environments. This work has compelling potential for increasing the environmental suitability of intelligent agents, such as drones, autonomous vehicles, and underwater robots.

Multidomain Adaptation With Sample and Source Distillation.

Unsupervised multidomain adaptation attracts increasing attention as it delivers richer information when tackling a target task from an unlabeled target domain by leveraging the knowledge attained from labeled source domains. However, it is the quality of training samples, not just the quantity, that influences transfer performance. In this article, we propose a multidomain adaptation method with sample and source distillation (SSD), which develops a two-step selective strategy to distill source samples and define the importance of source domains. To distill samples, the pseudo-labeled target domain is constructed to learn a series of category classifiers to identify transfer and inefficient source samples. To rank domains, the agreements of accepting a target sample as the insider of source domains are estimated by constructing a domain discriminator based on selected transfer source samples. Using the selected samples and ranked domains, transfer from source domains to the target domain is achieved by adapting multilevel distributions in a latent feature space. Furthermore, to explore more usable target information which is expected to enhance the performance across domains of source predictors, an enhancement mechanism is built by matching selected pseudo-labeled and unlabeled target samples. The degrees of acceptance learned by the domain discriminator are finally employed as source merging weights to predict the target task. Superiority of the proposed SSD is validated on real-world visual classification tasks.

An Extremely Simple Algorithm for Source Domain Reconstruction.

The aim of unsupervised domain adaptation (UDA) is to utilize knowledge from a source domain to enhance the performance of a given target domain. Due to the lack of accessibility to the target domain's labels, UDA's efficacy is highly reliant on the source domain's quality. However, it is often impractical and expensive to obtain an appropriate transferable source domain. To address this issue, we propose a novel UDA setting, source domain reconstruction (SDR), which seeks to construct a new transferable source domain utilizing labeled source samples and unlabeled target samples. SDR has a significant advantage over the conventional method as it is much less expensive to construct a suitable pseudo-source domain rather than collecting an actual transferable source domain in real-world scenarios. To test the practice of SDR, we investigate SDR theoretically. We propose an easily implementable algorithm, the domain MixUp (DMU), which is motivated by the MixUp strategy, to solve the SDR problem. The algorithm can be used to design a UDA framework to significantly enhance the performance of several existing UDA algorithms. Results from extensive experiments conducted on seven benchmarks (66 UDA tasks) indicate that the reconstructed source domain has stronger transferability than the original source domain.

From imaging to clinical outcome: dual-region CT radiomics predicting FOXM1 expression and prognosis in hepatocellular carcinoma.

Background: Liver cancer, especially hepatocellular carcinoma (HCC), remains a significant global health challenge. Traditional prognostic indicators for HCC often fall short in providing comprehensive insights for individualized treatment. The integration of genomics and radiomics offers a promising avenue for enhancing the precision of HCC diagnosis and prognosis. Methods: From the Cancer Genome Atlas (TCGA) database, we categorized mRNA of HCC patients by Forkhead Box M1 (FOXM1) expression and performed univariate and multivariate studies to pinpoint autonomous HCC risk factors. We deployed subgroup, correlation, and interaction analyses to probe FOXM1's link with clinicopathological elements. The connection between FOXM1 and immune cells was evaluated using the CIBERSORTx database. The functions of FOXM1 were investigated through analyses of Gene Ontology (GO) and the Kyoto Encyclopedia of Genes and Genomes (KEGG). After filtering through TCGA and the Cancer Imaging Archive (TCIA) database, we employed dual-region computed tomography (CT) radiomics technology to noninvasively predict the mRNA expression of FOXM1 in HCC tissues. Radiomic features were extracted from both tumoral and peritumoral regions, and a radiomics score (RS) was derived. The performance and robustness of the constructed models were evaluated using 10-fold cross-validation. A radiomics nomogram was developed by incorporating RS and clinical variables from the TCGA database. The models' discriminative abilities were assessed using metrics such as the area under the curve (AUC) of the receiver operating characteristic curves (ROC) and precision-recall (PR) curves. Results: Our findings emphasized the overexpression of FOXM1 as a determinant of poor prognosis in HCC and illustrated its impact on immune cell infiltration. After selecting arterial phase CT, we chose 7 whole-tumor features and 3 features covering both the tumor and its surroundings to create WT and WP models for FOXM1 prediction. The WT model showed strong predictive capabilities for FOXM1 expression by PR curve. Conversely, the WP model did not demonstrate the good predictive ability. In our study, the radiomics score (RS) was derived from whole-tumor regions on CT images. The RS was significantly associated with FOXM1 expression, with an AUC of 0.918 in the training cohort and 0.837 in the validation cohort. Furthermore, the RS was correlated with oxidative stress genes and was integrated with clinical variables to develop a nomogram, which demonstrated good calibration and discrimination in predicting 12-, 36-, and 60-month survival probabilities. Additionally, bioinformatics analysis revealed FOXM1's potential role in shaping the immune microenvironment, with its expression linked to immune cell infiltration. Conclusion: This study highlights the potential of integrating FOXM1 expression and radiomics in understanding HCC's complexity. Our approach offers a new perspective in utilizing radiomics for non-invasive tumor characterization and suggests its potential in providing insights into molecular profiles. Further research is needed to validate these findings and explore their clinical implications in HCC management.

Disentangling Stochastic PDE Dynamics for Unsupervised Video Prediction.

Unsupervised video prediction aims to predict future outcomes based on the observed video frames, thus removing the need for supervisory annotations. This research task has been argued as a key component of intelligent decision-making systems, as it presents the potential capacities of modeling the underlying patterns of videos. Essentially, the challenge of video prediction is to effectively model the complex spatiotemporal and often uncertain dynamics of high-dimensional video data. In this context, an appealing way of modeling spatiotemporal dynamics is to explore prior physical knowledge, such as partial differential equations (PDEs). In this article, considering real-world video data as a partly observed stochastic environment, we introduce a new stochastic PDE predictor (SPDE-predictor), which models the spatiotemporal dynamics by approximating a generalized form of PDEs while dealing with the stochasticity. A second contribution is that we disentangle the high-dimensional video prediction into low-level dimensional factors of variations: time-varying stochastic PDE dynamics and time-invariant content factors. Extensive experiments on four various video datasets show that SPDE video prediction model (SPDE-VP) outperforms both deterministic and stochastic state-of-the-art methods. Ablation studies highlight our superiority driven by both PDE dynamics modeling and disentangled representation learning and their relevance in long-term video prediction.

Fast treatment and recycling method of large-scale vegetable wastes.

With rapid development of vegetable industry in China, in process of refrigerated transportation and storage, large-scale abandoned vegetable wastes (VW) need to be urgently treated alone since they rot very fast and would pollute the environment seriously. Existing treatment projects generally regard VW as garbage with high content of water and adopt the process of squeeze and sewage treatment, which leads to not only high treatment costs but also great resource waste. Therefore, according to the composition and degradation characteristics of VW, a novel fast treatment and recycling method of VW was proposed in this paper. VW are first degraded with thermostatic anaerobic digestion (AD) and then the residues decompose rapidly with thermostatic aerobic digestion to meet the farmland application standard. To verify the feasibility of the method, the pressed VW water (PVW) and VW from the VW treatment plant were mixed and degraded in two 0.56 m3 digesters, and degraded substances were continuously measured in 30 days' mesophilic AD at 37 ± 1 °C. Subsequently, the biogas slurry (BS) produced by AD is decomposed by thermostatic aerobic aeration decomposition at 30 °C for 48 h to rapidly decompose. BS was confirmed to use safely for plants by germination index (GI) test. The results show that 96 % chemical oxygen demand (COD) from 15,711 mg/L to 1000 mg/L within 31 days and the GI of treated BS was 81.75 %. Besides, nutrient elements of N, P, and K keep good abundance, no heavy metals, pesticide residue, and hazardous substances were found. Other parameters were all lower than the BS placed for a half-year. VW are fast-treated and recycled with the new method, which provides a novel method for fast treatment and recycling of large-scale VW.

Nanoparticles Mediated circROBO1 Silencing to Inhibit Hepatocellular Carcinoma Progression by Modulating miR-130a-5p/CCNT2 Axis.

Background: Circular RNAs (circRNAs) are becoming vital biomarkers and therapeutic targets for malignant tumors due to their high stability and specificity in tissues. However, biological functions of circRNAs in hepatocellular carcinoma (HCC) are still not well studied. Methods: Gene Expression Omnibus (GEO) database and qRT-PCR were used to evaluate expression of circROBO1 (hsa_circ_0066568) in HCC tissues and cell lines. CCK-8, colony formation, EdU staining, flow cytometry for cell cycle analysis, and xenograft model assays were performed to detect the circROBO1 function in vitro and in vivo. RNA pull-down, RNA immunoprecipitation (RIP), and Luciferase reporter assays were used to investigate the relationship among circROBO1, miR-130a-5p, and CCNT2. More importantly, we developed nanoparticles made from poly lactic-co-glycolic acid (PLGA) and polyethylene glycol (PEG) chains as the delivery system of si-circROBO1 and then applied them to HCC in vitro and in mice. Results: circROBO1 was obviously upregulated in HCC tissues and cell lines, and elevated circROBO1 was closely correlated with worse prognosis for HCC patients. Functionally, knocking down circROBO1 significantly suppressed HCC cells growth in vitro and in mice. Mechanistically, circROBO1 acted as a competing endogenous RNA to downregulate miR-130a-5p, leading to CCNT2 expression upregulation. Furthermore, miR-130a-5p mimic or CCNT2 knockdown reversed the role of circROBO1 overexpression on HCC cells, which demonstrated that circROBO1 promoted HCC development via miR-130a-5p/CCNT2 axis. In addition, we developed nanoparticles loaded with si-circROBO1, named as PLGA-PEG (si-circROBO1) NPs, which significantly prevented the proliferation of HCC cells, and did not exhibit apparent toxicity to major organs in vivo. Conclusion: Our findings firstly demonstrate that circROBO1 overexpression promotes HCC progression by regulating miR-130a-5p/CCNT2 axis, which may serve as an effective nanotherapeutic target for HCC treatment.

A Novel Anti-Flashover Superhydrophobic Coating with Self-Assembly Characteristic of Surface Energy Differences.

Because of the versatility of superhydrophobic materials, they have attracted a lot of attention even in power electronics, transportation, engineering, and other fields. The volume fraction of fluorinated silicon oxide nanoparticles in superhydrophobic materials is one of the most important factors. Increasing the volume fraction will decrease the stability between the coating and the hydrophobic surface. Especially, the flashover voltage of the coating gradually decreases from 10 to 35 vol.%. Meanwhile, the flashover voltage dispersion of the coating increases drastically after 30 vol.%. In order to improve the electrical properties of the superhydrophobic coating, self-assembly of surface energy differences strategy is proposed in this work. A binary filling phase of the coating is introduced by 2D boron nitride nanosheets and silicon oxide nanoparticles. Although Hexagonal boron nitride with high surface energy and low roughness, it will be spontaneously assembled and wrapped by silicon oxide nanoparticle based on surface energy differences, which forming a low surface energy filled phase. Experiment results prove that the flashover voltage of the superhydrophobic coating is optimized by the binary filling phase coating. This method offers new ideas for the selection of filling phase and application of superhydrophobic materials.

Suppressing circIDE/miR-19b-3p/RBMS1 axis exhibits promoting-tumour activity through upregulating GPX4 to diminish ferroptosis in hepatocellular carcinoma.

Ferroptosis is a newly characterized form of iron-dependent non-apoptotic cell death, which is closely associated with cancer progression. However, the functions and mechanisms in regulation of escaping from ferroptosis during hepatocellular carcinoma (HCC) progression remain unknown. In this study, we reported that the RNA binding motif single stranded interacting protein 1 (RBMS1) participated in HCC development，and functioned as a regulator of ferroptosis. Clinically, the downregulation of RBMS1 occurred in HCC tissues, and low RBMS1 expression was associated with worse HCC patients survival. Mechanistically, RBMS1 overexpression inhibited HCC cell growth by attenuating the expression of glutathione peroxidase 4 (GPX4)and further facilitated ferroptosis in vitro and in vivo. More importantly, a novel circIDE (hsa_circ_0000251) was identified to elevate RBMS1 expression via sponging miR-19b-3p in HCC cells. Collectively, our findings established circIDE/miR-19b-3p/RBMS1 axis as a regulator of ferroptosis, which could be a promising therapeutic target and prognostic factor.

Learning Data Streams With Changing Distributions and Temporal Dependency.

In a data stream, concept drift refers to unpredictable distribution changes over time, which violates the identical-distribution assumption required by conventional machine learning methods. Current concept drift adaptation techniques mostly focus on a data stream with changing distributions. However, since each variable of a data stream is a time series, these variables normally have temporal dependency problems in the real world. How to solve concept drift and temporal dependency problems at the same time is rarely discussed in the concept-drift literature. To solve this situation, this article proves and validates that the testing error decreases faster if a predictor is trained on a temporally reconstructed space when drift occurs. Based on this theory, a novel drift adaptation regression (DAR) framework is designed to predict the label variable for data streams with concept drift and temporal dependency. A new statistic called local drift degree (LDD+) is proposed and used as a drift adaptation technique in the DAR framework to discard outdated instances in a timely way, thereby guaranteeing that the most relevant instances will be selected during the training process. The performance of DAR is demonstrated by a set of experimental evaluations on both synthetic data and real-world data streams.

Heterogeneous Multidomain Recommender System Through Adversarial Learning.

To solve the user data sparsity problem, which is the main issue in generating user preference prediction, cross-domain recommender systems transfer knowledge from one source domain with dense data to assist recommendation tasks in the target domain with sparse data. However, data are usually sparsely scattered in multiple possible source domains, and in each domain (source/target) the data may be heterogeneous, thus it is difficult for existing cross-domain recommender systems to find one source domain with dense data from multiple domains. In this way, they fail to deal with data sparsity problems in the target domain and cannot provide an accurate recommendation. In this article, we propose a novel multidomain recommender system (called HMRec) to deal with two challenging issues: 1) how to exploit valuable information from multiple source domains when no single source domain is sufficient and 2) how to ensure positive transfer from heterogeneous data in source domains with different feature spaces. In HMRec, domain-shared and domain-specific features are extracted to enable the knowledge transfer between multiple heterogeneous source and target domains. To ensure positive transfer, the domain-shared subspaces from multiple domains are maximally matched by a multiclass domain discriminator in an adversarial learning process. The recommendation in the target domain is completed by a matrix factorization module with aligned latent features from both the user and the item side. Extensive experiments on four cross-domain recommendation tasks with real-world datasets demonstrate that HMRec can effectively transfer knowledge from multiple heterogeneous domains collaboratively to increase the rating prediction accuracy in the target domain and significantly outperforms six state-of-the-art non-transfer or cross-domain baselines.

Bridging the Theoretical Bound and Deep Algorithms for Open Set Domain Adaptation.

In the unsupervised open set domain adaptation (UOSDA), the target domain contains unknown classes that are not observed in the source domain. Researchers in this area aim to train a classifier to accurately: 1) recognize unknown target data (data with unknown classes) and 2) classify other target data. To achieve this aim, a previous study has proven an upper bound of the target-domain risk, and the open set difference, as an important term in the upper bound, is used to measure the risk on unknown target data. By minimizing the upper bound, a shallow classifier can be trained to achieve the aim. However, if the classifier is very flexible [e.g., deep neural networks (DNNs)], the open set difference will converge to a negative value when minimizing the upper bound, which causes an issue where most target data are recognized as unknown data. To address this issue, we propose a new upper bound of target-domain risk for UOSDA, which includes four terms: source-domain risk, ϵ -open set difference ( ∆ϵ ), distributional discrepancy between domains, and a constant. Compared with the open set difference, ∆ϵ is more robust against the issue when it is being minimized, and thus we are able to use very flexible classifiers (i.e., DNNs). Then, we propose a new principle-guided deep UOSDA method that trains DNNs via minimizing the new upper bound. Specifically, source-domain risk and ∆ϵ are minimized by gradient descent, and the distributional discrepancy is minimized via a novel open set conditional adversarial training strategy. Finally, compared with the existing shallow and deep UOSDA methods, our method shows the state-of-the-art performance on several benchmark datasets, including digit recognition [modified National Institute of Standards and Technology database (MNIST), the Street View House Number (SVHN), U.S. Postal Service (USPS)], object recognition (Office-31, Office-Home), and face recognition [pose, illumination, and expression (PIE)].

Evolving Gradient Boost: A Pruning Scheme Based on Loss Improvement Ratio for Learning Under Concept Drift.

In nonstationary environments, data distributions can change over time. This phenomenon is known as concept drift, and the related models need to adapt if they are to remain accurate. With gradient boosting (GB) ensemble models, selecting which weak learners to keep/prune to maintain model accuracy under concept drift is nontrivial research. Unlike existing models such as AdaBoost, which can directly compare weak learners' performance by their accuracy (a metric between [0, 1]), in GB, weak learners' performance is measured with different scales. To address the performance measurement scaling issue, we propose a novel criterion to evaluate weak learners in GB models, called the loss improvement ratio (LIR). Based on LIR, we develop two pruning strategies: 1) naive pruning (NP), which simply deletes all learners with increasing loss and 2) statistical pruning (SP), which removes learners if their loss increase meets a significance threshold. We also devise a scheme to dynamically switch between NP and SP to achieve the best performance. We implement the scheme as a concept drift learning algorithm, called evolving gradient boost (LIR-eGB). On average, LIR-eGB delivered the best performance against state-of-the-art methods on both stationary and nonstationary data.

Semi-Supervised Heterogeneous Domain Adaptation: Theory and Algorithms.

Semi-supervised heterogeneous domain adaptation (SsHeDA) aims to train a classifier for the target domain, in which only unlabeled and a small number of labeled data are available. This is done by leveraging knowledge acquired from a heterogeneous source domain. From algorithmic perspectives, several methods have been proposed to solve the SsHeDA problem; yet there is still no theoretical foundation to explain the nature of the SsHeDA problem or to guide new and better solutions. Motivated by compatibility condition in semi-supervised probably approximately correct (PAC) theory, we explain the SsHeDA problem by proving its generalization error - that is, why labeled heterogeneous source data and unlabeled target data help to reduce the target risk. Guided by our theory, we devise two algorithms as proof of concept. One, kernel heterogeneous domain alignment (KHDA), is a kernel-based algorithm; the other, joint mean embedding alignment (JMEA), is a neural network-based algorithm. When a dataset is small, KHDA's training time is less than JMEA's. When a dataset is large, JMEA is more accurate in the target domain. Comprehensive experiments with image/text classification tasks show KHDA to be the most accurate among all non-neural network baselines, and JMEA to be the most accurate among all baselines.

[Accumulation Characteristics and Sources of PAEs in Agricultural Soils in Gansu Province].

In order to understand the composition and accumulation characteristics of phthalates esters (PAEs) in agricultural soils in Gansu province, a total of 41 soil samples from four agricultural soils in Gansu province were collected, and the content of six PAEs compounds was analyzed using a gas chromatography-single quadrupole mass spectrometer (GC-MS). The results showed that the average value of PAEs compounds in agricultural soils in Gansu province was 432.4 µg·kg-1. The detection rates of DMP, DEP, DnBP, DEHP, and DNOP in the soil were 100%, and BBP was not detected. The order of the average value of PAEs content in the four agricultural soils in Gansu province was:greenhouse>farmland (open field)>forest>grassland. The exceeding rates of dibutyl phthalate (DnBP), dimethyl phthalate (DMP), and dimethyl phthalate (DEP) were 94%, 28%, and 27%, and the remaining three did not exceed the standard. The composition of PAEs in different agricultural soils was different due to their different sources. DEHP and DnBP components in the six different PAEs monomers accounted for a higher proportion and were the main pollutants of PAEs in agricultural soils in Gansu province. In this study, the contents of soil PAEs and DEHP were significantly positively correlated with the residual amount of mulch film in the farmland (P<0.05). In general, the content of soil PAEs in the Hexi area of Gansu province was significantly higher than that in the Longdong area.

Robust Gaussian Process Regression With Input Uncertainty: A PAC-Bayes Perspective.

The Gaussian process (GP) algorithm is considered as a powerful nonparametric-learning approach, which can provide uncertainty measurements on the predictions. The standard GP requires clearly observed data, unexpected perturbations in the input may lead to learned regression model mismatching. Besides, GP also suffers from the lack of good generalization performance guarantees. To deal with data uncertainty and provide a numerical generalization performance guarantee on the unknown data distribution, this article proposes a novel robust noisy input GP (NIGP) algorithm based on the probably approximately correct (PAC) Bayes theory. Furthermore, to reduce the computational complexity, we develop a sparse NIGP algorithm, and then develop a sparse PAC-Bayes NIGP approach. Compared with NIGP algorithms, instead of maximizing the marginal log likelihood, one can optimize the PAC-Bayes bound to pursue a tighter generalization error upper bound. Experiments verify that the NIGP algorithms can attain greater accuracy. Besides, the PAC-NIGP algorithms proposed herein can achieve both robust performance and improved generalization error upper bound in the face of both uncertain input and output data.

The anti-alcoholism drug disulfiram effectively ameliorates ulcerative colitis through suppressing oxidative stresses-associated pyroptotic cell death and cellular inflammation in colonic cells.

BACKGROUND: Oxidative stress, cell pyroptosis and inflammation are considered as important pathogenic factors for ulcerative colitis (UC) development, and the traditional anti-alcoholism drug disulfiram (DSF) has recently been reported to exert its regulating effects on all the above cellular functions, which makes DSF as ideal therapeutic agent for UC treatment, but this issue has not been fully studied. METHODS: Dextran sulfate sodium (DSS)-induced animal models in C57BL/6J mice and lipopolysaccharide (LPS)-induced cellular models in colonic cell lines (HT-29 and Caco-2) for UC were respectively established. Cytokine secretion was determined by ELISA. Cell viability and proliferation were evaluated by MTT assay and EdU assay. Real-Time qPCR, Western Blot, immunofluorescent staining assay and immunohistochemistry (IHC) were employed to evaluate gene expressions. The correlations of the genes in the clinical tissues were analyzed by using the Pearson Correlation analysis. RESULTS: DSF restrained oxidative stress, pyroptotic cell death and cellular inflammation in UC models in vitro and in vivo, and elimination of Reactive Oxygen Species (ROS) by N-acetyl-l-cysteine (NAC) rescued cell viability in LPS-treated colonic cells (HT-29 and Caco-2). Further experiments suggested that a glycogen synthase kinase-3ß (GSK-3ß)/Nrf2/NLRP3 signaling cascade played critical role in this process. Mechanistically, DSF downregulated GSK-3ß and NLRP3, whereas upregulated Nrf2 in LPS-treated colonic cells. Also, the regulating effects of DSF on Nrf2 and NLRP3 were abrogated by upregulating GSK-3ß. Moreover, upregulation of GSK-3ß abolished the protective effects of DSF on LPS-treated colonic cells. CONCLUSIONS: Taken together, data of this study indicated that DSF restrained oxidative damages-related pyroptotic cell death and inflammation via regulating the GSK-3ß/Nrf2/NLRP3 pathway, leading to the suppression of LPS-induced UC development.

Multiclass Classification With Fuzzy-Feature Observations: Theory and Algorithms.

The theoretical analysis of multiclass classification has proved that the existing multiclass classification methods can train a classifier with high classification accuracy on the test set, when the instances are precise in the training and test sets with same distribution and enough instances can be collected in the training set. However, one limitation with multiclass classification has not been solved: how to improve the classification accuracy of multiclass classification problems when only imprecise observations are available. Hence, in this article, we propose a novel framework to address a new realistic problem called multiclass classification with imprecise observations (MCIMO), where we need to train a classifier with fuzzy-feature observations. First, we give the theoretical analysis of the MCIMO problem based on fuzzy Rademacher complexity. Then, two practical algorithms based on support vector machine and neural networks are constructed to solve the proposed new problem. The experiments on both synthetic and real-world datasets verify the rationality of our theoretical analysis and the efficacy of the proposed algorithms.

Metabolic Profiling Analysis of the Effect and Mechanism of Gushiling Capsule in Rabbits With Glucocorticoid-Induced Osteonecrosis of the Femoral Head.

Gushiling capsule (GSLC) is an effective traditional Chinese medicine for the treatment of glucocorticoid-induced osteonecrosis of the femoral head (GIONFH). This study established the serum metabolite profiles of GSLC in rabbits and explored the metabolic mechanism and effect of GSLC on GIONFH. Seventy-five Japanese white rabbits were randomly divided into the control, model, and GSLC groups. The rabbits in the model group and the GSLC group received injection of prednisolone acetate. Meanwhile, rabbits in the GSLC group were treated by gavage at a therapeutic dose of GSLC once a day. The control group and the model group received the same volume of normal saline gavage. Three groups of serum samples were collected at different time points, and the changes in the metabolic spectrum were analyzed by ultra-high performance liquid chromatography-tandem mass spectrometry (UPLC-MS/MS). The resulting data set was analyzed using multivariate statistical analysis to identify potential biomarkers related to GSLC treatment. The metabolic pathway was analyzed by MetaboAnalyst 4.0 and a heatmap was constructed using the HEML1.0.3.7 software package. In addition, histopathological and radiography studies were carried out to verify the anti-GIONFH effects of GSLC. Principal component analysis (PCA) and partial least squares-discriminant analysis (PLS-DA) score plots revealed a significant separation trend between the control group and the model group and the GSLC group (1-3 weeks), but there were no significant differences in the GSLC group (4-6 weeks). Orthogonal PLS-DA (OPLS-DA) score plots also revealed an obvious difference between the model and the GSLC groups (4-6 weeks). Ten potential metabolite biomarkers, mainly phospholipids, were identified in rabbit serum samples and demonstrated to be associated with GIONFH. Hematoxylin and eosin staining and magnetic resonance imaging indicated that the pathological changes in femoral head necrosis in the GSLC group were less than in the model group, which was consistent with the improved serum metabolite spectrum. GSLC regulated the metabolic disorder of endogenous lipid components in GIONFH rabbits. GSLC may prevent and treat GIONFH mainly by regulating phospholipid metabolism in vivo.

Screening of Potential Key Biomarkers for Ewing Sarcoma: Evidence from Gene Array Analysis.

Background: Ewing's sarcoma (ES) is a common bone cancer in children and adolescents. There are ethnic differences in the incidence and treatment effects. People have made great efforts to clarify the cause; however, the molecular mechanism of ES is still poorly understood. Methods: We download the microarray datasets GSE68776, GSE45544 and GSE17674 from the Gene Expression Omnibus (GEO) database. The differentially expressed genes (DEGs) of the three datasets were screened and enrichment analysis was performed. STRING and Cytoscape were used to carry out module analysis, building a protein-protein interaction (PPI) network. Finally, a series of analyses such as survival analysis and immune infiltration analysis were performed on the selected genes. Results: A total of 629 differentially expressed genes were screened, including 206 up-regulated genes and 423 down-regulated genes. The pathways and rich-functions of DEGs include protein activation cascade, carbohydrate binding, cell-cell adhesion junctions, mitotic cell cycle, p53 pathway, and cancer pathways. Then, a total of 10 hub genes were screened out. Biological process analysis showed that these genes were mainly enriched in mitotic nuclear division, protein kinase activity, cell division, cell cycle, and protein phosphorylation. Conclusion: Survival analysis and multiple gene comparison analysis showed that CDCA8, MAD2L1 and FANCI may be involved in the occurrence and prognosis of ES. The purpose of our study is to clarify the DEG and key genes, which will help us know more about the molecular mechanisms of ES, provide potential pathway or targets for the diagnosis and treatment.