Construction of a Prediction Model for Post-Thrombotic Syndrome After Deep Vein Thrombosis Incorporating Novel Inflammatory Response Parameter Scoring.

OBJECTIVE: To investigate the independent predictive factors for post-thrombotic syndrome (PTS) and to construct a risk prediction model for PTS by incorporating a novel inflammatory response parameter scoring. METHODS: A retrospective study analyzed patients diagnosed with lower extremity deep vein thrombosis (LEDVT) at the Affiliated Hospital of Chengde Medical College from January 2018 to January 2022. The Villalta scale was used to assess the occurrence of PTS 6-24 months after discharge. Patients were randomly divided into a training set and a validation set at a ratio of 7:3. In the training set, univariate analysis was performed on meaningful continuous variables, and those with differences were converted into dichotomous variables based on optimal cutoff values. Variable selection was performed using Log-Lambda and LASSO 10-fold cross-validation, followed by multivariable logistic regression analysis on selected variables for model construction. The model underwent internal validation in the validation set and external validation in an independent external cohort, including discriminative analysis, calibration analysis, and clinical decision curve analysis, with the model's rationale being evaluated lastly. RESULTS: A total of 356 patients with lower extremity DVT were included, with 249 in the training set for model construction and 107 in the validation set for internal validation, along with 37 external patients for external validation. A composite score of inflammatory response parameters, including the neutrophil to lymphocyte ratio (NLR), platelet to lymphocyte ratio (PLR), and monocyte to high-density lipoprotein cholesterol ratio (MHR) (NLR-PLR-MHR score, NPMscore), was developed, showing a significantly higher NPMscore in the PTS group compared to the non-PTS group (p<0.05). Predictive factors related to the risk of PTS occurrence included stage (OR=6.83, 95%CI: 2.74-18.04), varicose veins (OR=7.30, 95%CI: 2.29-25.75), homocysteine (Hcy) (OR=1.12, 95%CI: 1.04-1.22), NPMscore (OR=3.13, 95%CI: 1.94-5.36), standardized anticoagulant therapy (OR=5.77, 95%CI: 1.25-27.62), and one-stop treatment (OR=0.04, 95%CI: 0.00-0.35) were incorporated into the Nomogram model. The model showed good discrimination with a concordance index of 0.918 (95%CI: 0.876-0.959) for model construction, 0.843 (95%CI: 0.741-0.945) for internal validation, and 0.823 (95%CI: 0.667-0.903) for external validation. The Nomogram model, internal and external validation calibration curves showed good agreement between observed and predicted values. Decision curve analysis (DCA) indicated the Nomogram model predicted PTS risk probability thresholds ranging from 3%-98% for model construction, 5%-97% for internal validation, and 10%-80% for external validation, demonstrating better net benefit for predicting PTS risk in the model, internal, and external validation. Rationality analysis showed the model and internal validation had higher discrimination and clinical net benefit than other clinical indices. CONCLUSION: The novel inflammatory response parameter score (NPMscore) combined with stage, varicose veins, homocysteine (Hcy), standardized anticoagulant therapy, and one-stop treatment in the Nomogram model provides a practical tool for healthcare professionals to assess the risk of PTS in DVT patients, enabling early identification of high-risk patients for effective PTS prevention.

Attention-based investigation and solution to the trade-off issue of adversarial training.

Adversarial training has become the mainstream method to boost adversarial robustness of deep models. However, it often suffers from the trade-off dilemma, where the use of adversarial examples hurts the standard generalization of models on natural data. To study this phenomenon, we investigate it from the perspective of spatial attention. In brief, standard training typically encourages a model to conduct a comprehensive check to input space. But adversarial training often causes a model to overly concentrate on sparse spatial regions. This reduced tendency is beneficial to avoid adversarial accumulation but easily makes the model ignore abundant discriminative information, thereby resulting in weak generalization. To address this issue, this paper introduces an Attention-Enhanced Learning Framework (AELF) for robustness training. The main idea is to enable the model to inherit the attention pattern of standard pre-trained model through an embedding-level regularization. To be specific, given a teacher model built on natural examples, the embedding distribution of teacher model is used as a static constraint to regulate the embedding outputs of the objective model. This design is mainly supported with that the embedding feature of standard model is usually recognized as a rich semantic integration of input. For implementation, we present a simplified AELFs that can achieve the regularization with single cross entropy loss via the parameter initialization and parameter update strategy. This avoids the extra consistency comparison operation between embedding vectors. Experimental observations verify the rationality of our argument, and experimental results demonstrate that it can achieve remarkable improvements in generalization under the high-level robustness.

KIFC1 promotes proliferation and pseudo-bipolar division of ESCC through the transportation of Aurora B kinase.

Esophageal squamous cell carcinoma (ESCC) accounts for over 90% of total in China, and the five-year survival rate for patients is less than 30%. Accordingly, the identification of novel, effective early diagnosis markers and therapeutic targets for ESCC is of paramount importance. KIFC1 has been identified as highly expressed in several types of cancer, although its prognostic value is inconsistent, and no research has been conducted specifically on its effect on ESCC. To investigate the expression and function of KIFC1 in ESCC, we conducted immunohistochemical staining on 30 pairs of para-carcinoma tissue and cancerous tissues, revealing a significant increase in KIFC1 expression in ESCC tissues. Using siRNA to knock down KIFC1 significantly reduced the proliferation of EC109 ESCC cells both in vitro and in vivo. Bioinformatics analysis revealed a highly significant positive correlation between KIFC1 overexpression and signaling pathways associated with tumor proliferation pathways. In EC109 cells, overexpression of KIFC1 significantly increased the rate of centrosome amplification and the likelihood of pseudo-bipolar division. Furthermore, the expression of KIFC1 and the rate of centrosome amplification in ESCC tissues were also positively correlated. In order to explore the underline molecular mechanisms, we identified, through proteomics, that KIFC1 binds to the protein Aurora B. The knockdown of KIFC1 significantly reduced the distribution of Aurora B on the metaphase plate and substantially inhibited the phosphorylation of its classical substrate, Histone H3. In conclusion, these findings indicate the potential utility of KIFC1 as both a tumor marker and a promising target for therapeutic interventions.

Identification of KIFC1 as an independent prognostic marker in renal clear cell carcinoma correlates with tumor proliferation and immune infiltration.

Renal clear cell carcinoma (ccRCC) is the world's most common form of cancer. Up to a third will develop metastases; the 5-year survival rate of the patients was only 14%. Practical prognostic markers remain to be discovered. Kinesin-like protein (KIFC1), a critical factor in maintaining the stability of the microtubule system, has significant prognostic value in some tumors. We analyzed the prognostic value, associated signaling pathways, and regulatory mechanisms of KIFC1 in ccRCC through bioinformatics and proteomics. Concretely, both mRNA and protein expression levels of KIFC1 were dramatically upregulated. KIFC1 is an independent prognostic factor for ccRCC. The expression of KIFC1 showed a significant positive correlation (Spearman coefficient > 0.7) with tumor proliferation-related pathways (tumor proliferation, G2/M checkpoint, and DNA replication) and tumor inflammation. Further, intratumoral immune cell analysis revealed that high expression of KIFC1 predicted more infiltration of CD8 + T and CD4 + T cells (p < 0.001). However, there was a significant positive relationship between CD8 + T cells and numerous immune checkpoint genes. CD8 + T cells in tumors from the KIFC1 high expression group were at the dysregulated state. High expression of KIFC1 may predict a poor immunotherapy outcome. By proteomics, we analyzed proteins interacting with KIFC1; spliceosome proteins had the most significant enrichment, indicating the new directions for KIFC1 investigation. In conclusion, our study identified KIFC1 as an independent prognostic factor in renal clear cell carcinoma, and the associated processes involved tumor proliferation and immune infiltration. KIFC1 had a close relationship with spliceosome proteins; it may be a new research direction.

LibFewShot: A Comprehensive Library for Few-Shot Learning.

Few-shot learning, especially few-shot image classification, has received increasing attention and witnessed significant advances in recent years. Some recent studies implicitly show that many generic techniques or "tricks", such as data augmentation, pre-training, knowledge distillation, and self-supervision, may greatly boost the performance of a few-shot learning method. Moreover, different works may employ different software platforms, backbone architectures and input image sizes, making fair comparisons difficult and practitioners struggle with reproducibility. To address these situations, we propose a comprehensive library for few-shot learning (LibFewShot) by re-implementing eighteen state-of-the-art few-shot learning methods in a unified framework with the same single codebase in PyTorch. Furthermore, based on LibFewShot, we provide comprehensive evaluations on multiple benchmarks with various backbone architectures to evaluate common pitfalls and effects of different training tricks. In addition, with respect to the recent doubts on the necessity of meta- or episodic-training mechanism, our evaluation results confirm that such a mechanism is still necessary especially when combined with pre-training. We hope our work can not only lower the barriers for beginners to enter the area of few-shot learning but also elucidate the effects of nontrivial tricks to facilitate intrinsic research on few-shot learning.

Efficacy of Chinese medicine in the adjuvant treatment of Hashimoto's thyroiditis with hypothyroidism: a systematic review and meta-analysis.

The quality of the studies was assessed using the Cochrane risk bias assessment tool and the results were analyzed using Review Manager 5.3 software. The evidence was also evaluated for its strength using GRADE. A total of 18 randomized controlled trials were included in the study, involving 1247 patients. The primary outcomes of this study included overall efficacy, effectiveness in treating specific symptoms, and the Traditional Chinese Medicine symptom score. The secondary outcomes included the levels of FT3, FT4, and TSH, the size of the thyroid gland, and any adverse events. The results of the meta-analysis showed that CHM combined with WM has a better curative effect and a more effective reduction in clinical symptoms than WM alone: comprehensive efficacy [OR = 4.83; 95% CI (3.45, 6.76)], syndrome efficacy [OR = 5.95; 95% CI (3.94, 8.99)], TCM symptom score SMD = -1.49; 95% CI (-1.86, -1.11)], FT3 [SMD = 0.59; 95% CI (0.48, 0.71)], FT4 [SMD = 0.59; 95% CI (0.48, 0.71)], TSH SMD = -0.97; 95% CI (-1.35, -0.58)], and thyroid volume SMD = -0.25; 95% CI (-0.34, 0.15)]. The incidence of adverse events between the groups was not significantly different [OR = 1.00; 95% CI (0.14, 7.27)]. Because of the effectiveness of CHM, we support using CHM to improve clinical efficacy in the treatment of HTH. The results of our research suggest that the use of Chinese Herbal Medicine (CHM) in combination with Western Medicine (WM) may result in improved clinical efficacy in the treatment of hypothyroidism (HTH) compared to using WM alone.

Delayed Chronic Acidic Postconditioning Improves Poststroke Motor Functional Recovery and Brain Tissue Repair by Activating Proton-Sensing TDAG8.

Acidic postconditioning by transient CO2 inhalation applied within minutes after reperfusion has neuroprotective effects in the acute phase of stroke. However, the effects of delayed chronic acidic postconditioning (DCAPC) initiated during the subacute phase of stroke or other acute brain injuries are unknown. Mice received daily DCAPC by inhaling 5%/10%/20% CO2 for various durations (three cycles of 10- or 20-min CO2 inhalation/10-min break) at days 3-7, 7-21, or 3-21 after photothrombotic stroke. Grid-walk, cylinder, and gait tests were used to assess motor function. DCAPC with all CO2 concentrations significantly promoted motor functional recovery, even when DCAPC was delayed for 3-7 days. DCAPC enhanced the puncta density of GAP-43 (a marker of axon growth and regeneration) and synaptophysin (a marker of synaptogenesis) and reduced the amoeboid microglia number, glial scar thickness and mRNA expression of CD16 and CD32 (markers of proinflammatory M1 microglia) compared with those of the stroke group. Cerebral blood flow (CBF) increased in response to DCAPC. Furthermore, the mRNA expression of TDAG8 (a proton-activated G-protein-coupled receptor) was increased during the subacute phase of stroke, while DCAPC effects were blocked by systemic knockout of TDAG8, except for those on CBF. DCAPC reproduced the benefits by re-expressing TDAG8 in the peri-infarct cortex of TDAG8-/- mice infected with HBAAV2/9-CMV-TDAG8-3flag-ZsGreen. Taken together, we first showed that DCAPC promoted functional recovery and brain tissue repair after stroke with a wide therapeutic time window of at least 7 days after stroke. Brain-derived TDAG8 is a direct target of DCAPC that induces neuroreparative effects.

Net-zero transition of the global chemical industry with CO2-feedstock by 2050: feasible yet challenging.

Carbon capture, utilization and storage (CCUS) have been projected by the power and industrial sectors to play a vital role towards net-zero greenhouse gas emissions. In this study, we aim to explore the feasibility of a global chemical industry that fully relies on CO2 as its carbon source in 2050. We project the global annual CO2 demand as chemical feedstock to be 2.2-3.1 gigatonnes (Gt), well within the possible range of supply (5.2-13.9 Gt) from the power, cement, steel, and kraft pulp sectors. Hence, feedstock availability is not a constraint factor for the transition towards a fully CO2-based chemical industry on the global basis, with the exception of few regions that could face local supply shortages, such as the Middle East. We further conduct life cycle assessment to examine the environmental benefits on climate change and the trade-offs of particulate matter-related health impacts induced by carbon capture. We conclude that CO2 captured from solid biomass-fired power plants and kraft pulp mills in Europe would have the least environmental and health impacts, and that India and China should prioritize low-impact regional electricity supply before a large-scale deployment of CCUS. Finally, two bottom-up case studies of China and the Middle East illustrate how the total regional environmental and health impacts from carbon capture can be minimized by optimizing its supply sources and transport, requiring cross-sectoral cooperation and early planning of infrastructure. Overall, capture and utilization of unabatable industrial waste CO2 as chemical feedstock can be a feasible way for the net-zero transition of the industry, while concerted efforts are yet needed to build up the carbon-capture-and-utilization value chain around the world.

A Multilayer Framework for Online Metric Learning.

Online metric learning (OML) has been widely applied in classification and retrieval. It can automatically learn a suitable metric from data by restricting similar instances to be separated from dissimilar instances with a given margin. However, the existing OML algorithms have limited performance in real-world classifications, especially, when data distributions are complex. To this end, this article proposes a multilayer framework for OML to capture the nonlinear similarities among instances. Different from the traditional OML, which can only learn one metric space, the proposed multilayer OML (MLOML) takes an OML algorithm as a metric layer and learns multiple hierarchical metric spaces, where each metric layer follows a nonlinear layer for the complicated data distribution. Moreover, the forward propagation (FP) strategy and backward propagation (BP) strategy are employed to train the hierarchical metric layers. To build a metric layer of the proposed MLOML, a new Mahalanobis-based OML (MOML) algorithm is presented based on the passive-aggressive strategy and one-pass triplet construction strategy. Furthermore, in a progressively and nonlinearly learning way, MLOML has a stronger learning ability than traditional OML in the case of limited available training data. To make the learning process more explainable and theoretically guaranteed, theoretical analysis is provided. The proposed MLOML enjoys several nice properties, indeed learns a metric progressively, and performs better on the benchmark datasets. Extensive experiments with different settings have been conducted to verify these properties of the proposed MLOML.

Defensive Few-Shot Learning.

This article investigates a new challenging problem called defensive few-shot learning in order to learn a robust few-shot model against adversarial attacks. Simply applying the existing adversarial defense methods to few-shot learning cannot effectively solve this problem. This is because the commonly assumed sample-level distribution consistency between the training and test sets can no longer be met in the few-shot setting. To address this situation, we develop a general defensive few-shot learning (DFSL) framework to answer the following two key questions: (1) how to transfer adversarial defense knowledge from one sample distribution to another? (2) how to narrow the distribution gap between clean and adversarial examples under the few-shot setting? To answer the first question, we propose an episode-based adversarial training mechanism by assuming a task-level distribution consistency to better transfer the adversarial defense knowledge. As for the second question, within each few-shot task, we design two kinds of distribution consistency criteria to narrow the distribution gap between clean and adversarial examples from the feature-wise and prediction-wise perspectives, respectively. Extensive experiments demonstrate that the proposed framework can effectively make the existing few-shot models robust against adversarial attacks. Code is available at https://github.com/WenbinLee/DefensiveFSL.git.

Dietary ß-carotene and vitamin A and risk of Parkinson disease: A protocol for systematic review and meta-analysis.

BACKGROUND: The beneficial effects of dietary ß-carotene and vitamin A on Parkinson disease (PD) have been confirmed, but some studies have yielded questionable results. Therefore, this meta-analysis investigated the effect of dietary ß-carotene and vitamin A on the risk of PD. METHODS: The following databases were searched for relevant paper: PubMed, Embase, Medline, Scopus, Cochrane Library, CNKI, Wanfang Med online, and Weipu databases for the relevant paper from 1990 to March 28, 2022. The studies included were as follows: ß-carotene and vitamin A intake was measured using scientifically recognized approaches, such as food frequency questionnaire (FFQ); evaluation of odds ratios using OR, RR, or HR; ß-carotene and vitamin A intake for three or more quantitative categories; and PD diagnosed by a neurologist or hospital records. RESULTS: This study included 11 studies (four cohort studies, six case-control studies, and one cross-sectional study). The high ß-carotene intake was associated with a significantly lower chance of developing PD than low ß-carotene intake (pooled OR = 0.83, 95%CI = 0.74-0.94). Whereas the risk of advancement of PD was not significantly distinctive among the highest and lowest vitamin A intake (pooled OR = 1.08, 95%CI = 0.91-1.29). CONCLUSIONS: Dietary ß-carotene intake may have a protective effect against PD, whereas dietary vitamin A does not appear to have the same effect. More relevant studies are needed to include into meta-analysis in the further, as the recall bias and selection bias in retrospective and cross-sectional studies cause misclassifications in the assessment of nutrient intake.

Multislice Computed Tomography Angiography Imaging Diagnosis of Lower Extremity Arteriosclerosis in Patients with Hypertension and Its Correlation with the Level of High-Sensitivity C-Reactive Protein.

The aim of this study was to investigate the relationship between multislice computed tomography (CT) angiography (MSCTA) imaging and high-sensitivity C-reactive protein (hs-CRP) in patients with hypertension and lower extremity arteriosclerosis. 68 hypertensive patients with lower extremity arteriosclerosis were selected as the observation group, and 68 healthy volunteers were selected as the control group to compare the differences in hs-CRP. According to the degree of stenosis, the patients were further divided into five grades: no obvious stenosis, mild stenosis, moderate stenosis, severe stenosis, and occlusion. The correlation between the degree of stenosis and the content of hs-CRP was compared. The changes of hs-CRP content before and after treatment were compared, and the difference of images before and after surgical treatment and the difference of hs-CRP expression in patients with occlusion were compared. Compared with the control group, the content of hs-CRP in the observation group was significantly higher (P < 0.05), and the degree of stenosis was positively correlated with the content of hs-CRP. After two weeks of treatment, the hs-CRP levels of patients with severe stenosis and occlusion were significantly lower than those before treatment (P < 0.01). The level of hs-CRP in patients with occlusion after arterial stent intervention was significantly lower than before, and the images also showed that the blood vessels were significantly expanded. The degree of stenosis in patients with lower extremity arteriosclerosis diagnosed by MSCTA imaging was closely related to the expression of hs-CRP in the patient, and a sustained high concentration of hs-CRP corresponded to a more severe degree of vascular occlusion. In conclusion, the hs-CRP can be used as one of the factors to predict and evaluate the occurrence of cardiovascular and cerebrovascular diseases.

Toll-Like Receptor 4 Exacerbates Mycoplasma pneumoniaevia Promoting Transcription Factor EB-Mediated Autophagy.

Mycoplasma pneumoniae (M. pneumoniae) is the most common cause of community-acquired pneumonia. Toll-like receptors (TLRs) play an essential role in pneumonia. The purpose of this study was to investigate the roles of TLR4 in M. pneumoniae. Mice were administrated with 100 µl (1 × 107 ccu/ml) of M. pneumoniae. HE staining was applied for histological analysis. The protein expression was determined by western blot. The cytokine level was detected by ELISA. The results showed that TLR4-deficient mice were protected from M. pneumoniae. However, downregulation of TLR4 inhibited inflammatory response and autophagy. Moreover, transcription factor EB (TFEB) participated in M. pneumoniae-induced inflammatory response and autophagy, while knockdown of TLR4 downregulated TFEB and its nuclear translocation.

CAST: Learning Both Geometric and Texture Style Transfers for Effective Caricature Generation.

Given a photo of a subject, ability to generate a caricature image that captures distinct characteristics of the subject but with certain exaggeration of their prominent features is of fundamental importance to image processing and facial recognition. There are two main challenges in this task: shape exaggeration and style transfer. The former morphs and exaggerates key facial features of the subject, while the latter generates caricature images in a certain artistic style. In this paper, we propose a CAricature Style Transfer (CAST) framework for caricature generation. There are two modules in the proposed framework. The first is a geometric warping module. Different from the existing style transfer methods, we incorporate the Whitening and Coloring Transformation (WCT) in the geometric style transfer. The WCT is learned on photo and caricature landmarks or the caricature landmark space of a specific artist and is capable of transforming input photo landmarks to caricature landmarks. The second module is a texture style rendering module. We propose a new style transfer method by considering a semantic region-aligned style transfer via affinity constraint. Given a reference caricature image as the style reference, this module is capable of transferring styles between the same or similar semantic regions in caricatures and photos. Furthermore, it can transfer visual attributes of the reference caricatures (such as mouth shape and expressions) to the output caricatures. Experiments have shown desirable effects of the proposed method in transferring both the geometric and artistic texture styles of caricatures. Both qualitative and quantitative results show that the CAST framework is more effective compared than the state-of-the-art caricature generation methods.

Epi-illumination dark-field microscopy enables direct visualization of unlabeled small organisms with high spatial and temporal resolution.

Dark-field microscopy is known to offer both high resolution and direct visualization of thin samples. However, its performance and optimization on thick samples is under-explored and so far, only meso-scale information from whole organisms has been demonstrated. In this work, we carefully investigate the difference between trans- and epi-illumination configurations. Our findings suggest that the epi-illumination configuration is superior in both contrast and fidelity compared to trans-illumination, while having the added advantage of experimental simplicity and an "open top" for experimental intervention. Guided by the theoretical analysis, we constructed an epi-illumination dark-field microscope with measured lateral and axial resolutions of 260 nm and 520 nm, respectively. Subcellular structures in whole organisms were directly visualized without the need for image reconstruction, and further confirmed via simultaneous fluorescence imaging. With an imaging speed of 20 to 50 fps, we visualize fast dynamic processes such as cell division and pharyngeal pumping in Caenorhabditis elegans.

Detection of catecholamine metabolites in urine based on ultra-high-performance liquid chromatography-tandem mass spectrometry.

The excretion of neurotransmitter metabolites in normal individuals is of great significance for health monitoring. A rapid quantitative method was developed with ultra-performance liquid chromatography-tandem mass spectrometry. The method was further applied to determine catecholamine metabolites vanilymandelic acid (VMA), methoxy hydroxyphenyl glycol (MHPG), dihydroxy-phenyl acetic acid (DOPAC), and homovanillic acid (HVA) in the urine. The urine was collected from six healthy volunteers (20-22 years old) for 10 consecutive days. It was precolumn derivatized with dansyl chloride. Subsequently, the sample was analyzed using triple quadrupole mass spectrometry with an electrospray ion in positive and multireaction monitoring modes. The method was sensitive and repeatable with the recoveries 92.7-104.30%, limits of detection (LODs) 0.01-0.05 µg/mL, and coefficients no less than 0.9938. The excretion content of four target compounds in random urine samples was 0.20 ± 0.086 µg/mL (MHPG), 1.27 ± 1.24 µg/mL (VMA), 3.29 ± 1.36 µg/mL (HVA), and 1.13 ± 1.07 µg/mL (DOPAC). In the urine, the content of VMA, the metabolite of norepinephrine and adrenaline, was more than MHPG, and the content of HVA, the metabolite of dopamine, was more than DOPAC. This paper detected the levels of catecholamine metabolites and summarized the characteristics of excretion using random urine samples, which could provide valuable information for clinical practice.

Identification of Differentially Expressed Genes Particularly Associated with Immunity in Uremia Patients by Bioinformatic Analysis.

Uremia is a common syndrome that happens to nearly all end-stage kidney diseases, which profound have changes in human gene expressions, but the related pathways are poorly understood. Gene Ontology categories and Kyoto Encyclopedia of Genes and Genomes pathways enriched in the differentially expressed genes (DEGs) were analyzed by using clusterProfiler, org.Hs.eg.db, and Pathview, and protein-protein interaction (PPI) network was built by Cytoscape. We identified 3432 DEGs (including 3368 down- and 64 up-regulated genes), of which there were 52 different molecular functions, and 178 genes were identified as immune genes controlled by the four transcription factors (POU domain class 6 transcription factor 1 (POU6F1), interferon regulator factor 7 [IRF7], forkhead box D3 (FOXD3), and interferon-stimulated response element [ISRE]). In the gender research, no significant difference was observed. The top 15 proteins of 178 immune-related genes were identified with the highest degree in PPI network. The DEG analysis of uremia patients was expected to provide fundamental information to relieve pain and add years to their life.

MW-GAN: Multi-Warping GAN for Caricature Generation With Multi-Style Geometric Exaggeration.

Given an input face photo, the goal of caricature generation is to produce stylized, exaggerated caricatures that share the same identity as the photo. It requires simultaneous style transfer and shape exaggeration with rich diversity, and meanwhile preserving the identity of the input. To address this challenging problem, we propose a novel framework called Multi-Warping GAN (MW-GAN), including a style network and a geometric network that are designed to conduct style transfer and geometric exaggeration respectively. We bridge the gap between the style/landmark space and their corresponding latent code spaces by a dual way design, so as to generate caricatures with arbitrary styles and geometric exaggeration, which can be specified either through random sampling of latent code or from a given caricature sample. Besides, we apply identity preserving loss to both image space and landmark space, leading to a great improvement in quality of generated caricatures. Experiments show that caricatures generated by MW-GAN have better quality than existing methods.

A1/A2 astrocytes in central nervous system injuries and diseases: Angels or devils?

Astrocytes play a pivotal role in maintaining the central nervous system (CNS) homeostasis and function. In response to CNS injuries and diseases, reactive astrocytes are triggered. By purifying and genetically profiling reactive astrocytes, it has been now found that astrocytes can be activated into two polarization states: the neurotoxic or pro-inflammatory phenotype (A1) and the neuroprotective or anti-inflammatory phenotype (A2). Although the simple dichotomy of the A1/A2 phenotypes does not reflect the wide range of astrocytic phenotypes, it facilitates our understanding of the reactive state of astrocytes in various CNS disorders. This article reviews the recent evidences regarding A1/A2 astrocytes, including (a) the specific markers and morphological characteristics, (b) the effects of A1/A2 astrocytes on the neurovascular unit, and (c) the molecular mechanisms involved in the phenotypic switch of astrocytes. Although many questions remain, a deeper understanding of different phenotypic astrocytes will eventually help us to explore effective strategies for neurological disorders by targeting astrocytes.