Fusion Step Research Articles

Simultaneous Road Surface and Centerline Extraction From Large-Scale Remote Sensing Images Using CNN-Based Segmentation and Tracing

Accurate and up-to-date road maps are of great importance in a wide range of applications. Unfortunately, automatic road extraction from high-resolution remote sensing images remains challenging due to the occlusion of trees and buildings, discriminability of roads, and complex backgrounds. To address these problems, especially road connectivity and completeness, in this article, we introduce a novel deep learning-based multistage framework to accurately extract the road surface and road centerline simultaneously. Our framework consists of three steps: boosting segmentation, multiple starting points tracing, and fusion. The initial road surface segmentation is achieved with a fully convolutional network (FCN), after which another lighter FCN is applied several times to boost the accuracy and connectivity of the initial segmentation. In the multiple starting points tracing step, the starting points are automatically generated by extracting the road intersections of the segmentation results, which then are utilized to track consecutive and complete road networks through an iterative search strategy embedded in a convolutional neural network (CNN). The fusion step aggregates the semantic and topological information of road networks by combining the segmentation and tracing results to produce the final and refined road segmentation and centerline maps. We evaluated our method utilizing three data sets covering various road situations in more than 40 cities around the world. The results demonstrate the superior performance of our proposed framework. Specifically, our method’s performance exceeded the other methods by 7% and 40% for the connectivity indicator for road surface segmentation and for the completeness indicator for centerline extraction, respectively.

Chlorcyclizine Inhibits Viral Fusion of Hepatitis C Virus Entry by Directly Targeting HCV Envelope Glycoprotein 1

Chlorcyclizine (CCZ) is a potent hepatitis C virus (HCV) entry inhibitor, but its molecular mechanism is unknown. Here, we show that CCZ directly targets the fusion peptide of HCV E1 and interferes with the fusion process. Generation of CCZ resistance-associated substitutions of HCV invitro revealed six missense mutations in the HCV E1 protein, five being in the putative fusion peptide. A viral fusion assay demonstrated that CCZ blocked HCV entry at the membrane fusion step and that the mutant viruses acquired resistance to CCZ's action in blocking membrane fusion. UV cross-linking of photoactivatable CCZ-diazirine-biotin in both HCV-infected cells and recombinant HCV E1/E2 protein demonstrated direct binding to HCV E1 glycoprotein. Mass spectrometry analysis revealed that CCZ cross-linked to an E1 sequence adjacent to the putative fusion peptide. Docking simulations demonstrate a putative binding model, wherein CCZ binds to a hydrophobic pocket of HCV E1 and forms extensive interactions with the fusion peptide.

Health index extraction for power-shift steering transmission using selected oil field data

The concentration of wear debris in used lubricating oil is generally observed from spectral oil analysis. The element concentration data are one of the commonly used oil field data for degradation evaluation of power-shift steering transmission. However, in practical applications, the underlying degradation degree of the power-shift steering transmission is difficult to evaluate due to the complexity of element concentration data. To solve this problem, we propose a health index extraction methodology using a weighted average method to better characterize the underlying degradation, which leads to an accurate estimation of the residual life before the power-shift steering transmission needs to be maintained. The extracted health index not only maximizes the monotonic trend of underlying degradation but also minimizes the failure threshold variance between different power-shift steering transmissions. The method includes element concentration data modification, data selection, and data fusion steps that result in a reasonable power-shift steering transmission degradation model. Finally, a case study is provided to illustrate the proposed method. The results show that the extracted health index outperforms each selected element concentration data.

SCNET: A Novel UGI Cancer Screening Framework Based on Semantic-Level Multimodal Data Fusion.

Upper gastrointestinal (UGI) cancer has been identified as one of the ten most common causes of cancer deaths globally. UGI cancer screening is critical to improving the survival rate of UGI cancer patients. While many approaches to UGI cancer screening rely on single-modality data such as gastroscope imaging, limited studies have been dedicated to UGI cancer screening exploiting multisource and multimodal medical data, which could potentially lead to improved screening results. In this paper, we propose semantic-level cancer-screening network (SCNET), a framework for UGI cancer screening based on semantic-level multimodal upper gastrointestinal data fusion. Specifically, the proposed SCNET consists of a gastrointestinal image recognition flow and a textual medical record processing flow. High-level features of upper gastrointestinal data are extracted by identifying effective feature channels according to the correlation between the textual features and the spatial structure of the image features. The final screening results are obtained after the data fusion step. The experimental results show that the improvement of our approach over the state-of-the-art ones reached 4.01% in average. The source code of SCNET is available at https://github.com/netflymachine/SCNET.

Deep Pixel-Level Matching via Attention for Video Co-Segmentation

In video object co-segmentation, methods based on patch-level matching are widely leveraged to extract the similarity between video frames. However, these methods can easily lead to pixel misclassification because they reduce the precision of pixel localization; thus, the accuracies of the segmentation results of these methods are deducted. To address this problem, we propose a framework based on deep neural networks and equipped with a new attention module, which is designed for pixel-level matching to segment the object across video frames in this paper. In this attention module, the pixel-level matching step is able to compare the feature value of each pixel from one input frame with that of each pixel from another input frame for computing the similarity between two frames. Then a features fusion step is applied to efficiently fuse the feature maps of each frame with the similarity information for generating dense attention features. Finally, an up-sampling step refines the feature maps for obtaining high quality segmentation results by using these dense attention features. The ObMiC and DAVIS 2016 datasets were utilized to train and test our framework. Experimental results show that our framework achieves higher accuracy than those of other video segmentation methods that perform well in common information extraction.

Joint image deblurring and matching with feature-based sparse representation prior

Image matching aims to find a similar area of the small image in the large image, which is one of the key steps in image fusion and vision-based navigation; however, most matching methods perform poorly when the images to be matched are blurred. Traditional approaches for blurred image matching usually follow a two-stage framework - first resorting to image deblurring and then performing image matching with the recovered image. However, the matching accuracy of these methods often suffers greatly from the deficiency of image deblurring. Recently, a joint image deblurring and matching method that utilizes the sparse representation prior to exploit the correlation between deblurring and matching was proposed to address this problem and found to obtain a higher matching accuracy. Yet, that technique is not efficient when the image is seriously blurred, and the method’s time complexity is excessive. In this paper, we propose a joint image deblurring and matching approach with a feature-based sparse representation prior. Our approach utilizes two-directional two-dimensional (2D)2PCA to extract feature vectors from images and obtains a sparse representation prior in a robust feature space rather than the original pixel space, thus mitigating the influence of image blur. Moreover, the reduction in the feature dimension can also increase the computational efficiency. Extensive experiments show that our approach significantly outperforms state-of-the-art approaches in terms of both accuracy and speed.

Receptor-Interacting Protein Kinase 3 (RIPK3) inhibits autophagic flux during necroptosis in intestinal epithelial cells.

Autophagy is an intracellular process that regulates the degradation of cytosolic proteins and organelles. Dying cells often accumulate autophagosomes. However, the mechanisms by which necroptotic stimulation induces autophagosomes are not defined. Here, we demonstrate that the activation of necroptosis with TNF-α plus the cell-permeable pan-caspase inhibitor Z-VAD induces LC3-II and LC3 puncta, markers of autophagosomes, via the receptor-interacting protein kinase 3 (RIPK3) in intestinal epithelial cells. Surprisingly, necroptotic stimulation reduces autophagic activity, as evidenced by enlarged puncta of the autophagic substrate SQSTM1/p62 and its increased colocalization with LC3. However, necroptotic stimulation does not induce the lysosomal-associated membrane protein 1 (LAMP1) nor syntaxin 17, which mediates autophagosome-lysosome fusion, to colocalize with LC3. These data indicate that necroptosis attenuates autophagic flux before the lysosome fusion step. Our findings may provide insights into human diseases involving necroptosis.

Computational Modelling Framework to Study Ca2+ Activation of Synaptic Vesicle Fusion by Different Synaptotagmin Isoforms

Controlled release of neurotransmitters stored in synaptic vesicles (SVs) is central to information processing in the brain. This process relies on fast (sub-millisecond) coupling of SV fusion to the triggering signal - action potential (AP)-evoked presynaptic Ca2+ influx. During the last three decades the key proteins that mediate SV exocytosis have been identified. Yet,the molecular mechanisms that allow synchronisation of SV fusion to APs remain in the centre of debate. SV fusion is catalysed by synaptic SNARE proteins, VAMP2 on the vesicles (v-SNAREs) and syntaxin/SNAP25 on the pre-synaptic membrane (t-SNAREs). The current view is that the SNARE assembly process is arrested at a ‘half-zippered’ state to prevent continuous release of neurotransmitters and to allow for rapid and synchronous SV fusion following the Ca2+ influx. The key players that ‘clamp’ vesicle release and synchronise the late SV fusion steps are complexin proteins and isoforms of the presynaptic Ca2+ release sensor synaptotagmin (Syt). Here we describe a computational modelling framework that allows one to simulate AP-evoked synchronous and asynchronous release controlled by Syt1 and Syt7 isoforms under different conditions. The model is based on the recent structural data describing interaction of Syt1 with the SNARE complex and complexin. Ca2+ activation of SV fusion is simulated using the well-established Ca2+ and membrane-binding properties of Syt1 and Syt7 C2 domains. Conformational protein changes are represented by transitions between discrete states in a Markovian system, with parameters constrained by the known results for Ca2+ and membrane binding affinities. Our stochastic simulations demonstrate that this relatively simple model can successfully capture the kinetics and Ca2+ dependency of synchronous and asynchronous neurotransmitter release in central synapses.

Synaptic Vesicle Release Probability, Kinetics, and Ca-Sensitivity are Regulated by SNARE-Proteins

Release of neurotransmitters (NTs) at neuronal synapses is accomplished by a multi-component machinery that senses calcium influx into the presynaptic terminal and responds by fusing the synaptic vesicle and plasma membranes and releasing NTs through a fusion pore. Decades of studies assumed that the vesicle release probability, the vesicle release rate and the Ca-sensitivity are regulated by the Ca-sensing component of the machinery. Here, we use molecular dynamics simulations of the multi-protein release machinery to demonstrate that the membrane fusion component of the machinery, consisting of the SNARE proteins, is equally important for regulation: when more SNARE complexes are assembled at the fusion site, the release probability and Ca-sensitivity increase, demonstrating a remarkable coupling between the Ca-sensing and membrane-fusing submachineries. Our simulations access the long timescales of NT release by implementing systematically coarse-grained representations of the neuronal SNARE complexes (Mostafavi et al., 2017) and the Ca-sensor for synchronous NT release, Synaptotagmin 1. The model quantitatively explains experiments in which NT release was enhanced at mouse Calyx of Held synapses by a constitutively open mutant of the t-SNARE Syntaxin, which increased the number of assembled SNARE complexes at the fusion site (Acuna et al., 2014): the vesicle release probability, NT release rate and apparent Ca-sensitivity of release were all increased, while the cooperativity of release was unaffected. We find the increased release probability and Ca-sensitivity are caused by acceleration of the final membrane-fusion step due to increased entropic forces that push the membranes together, and coupling of this final fusion step with the first stage of the process in which calcium triggers unclamping of the fusion machinery. Thus, the NT release machinery consists of membrane-fusing and Ca-sensing submachineries that are intimately coupled and both contribute to regulation of release.

Secure Information Fusion using Local Posterior for Distributed Cyber-Physical Systems

In modern distributed cyber-physical systems (CPS), information fusion often plays a key role in automate and self-adaptive decision making process. However, given the heterogeneous and distributed nature of modern CPSs, it is a great challenge to operate CPSs with the compromised data integrity and unreliable communication links. In this paper, we study the distributed state estimation problem under the false data injection attack (FDIA) with probabilistic communication networks. We propose an integrated ”detection + fusion” solution, which is based on the Kullback-Leibler divergences (KLD) between local posteriors and therefore does not require the exchange of raw sensor data. For the FDIA detection step, the KLDs are used to cluster nodes in the probability space and to partition the space into secure and insecure subspaces. By approximating the distribution of the KLDs with a general <inline-formula><tex-math notation="LaTeX">$\chi ^2$</tex-math></inline-formula> distribution and calculating its tail probability, we provide an analysis of the detection error rate. For the information fusion step, we discuss the potential risk of double counting the shared prior information in the KLD-based consensus formulation method. We show that if the local posteriors are updated from the shared prior, the increased number of neighbouring nodes will lead to the diminished information gain. To overcome this problem, we propose a near-optimal distributed information fusion solution with properly weighted prior and data likelihood. Finally, we present simulation results for the integrated solution. We discuss the impact of network connectivity on the empirical detection error rate and the accuracy of state estimation.

An approach for feature selection with data modelling in LC-MS metabolomics.

The data processing workflow for LC-MS based metabolomics study is suggested with signal drift correction, univariate analysis, supervised learning, feature selection and unsupervised modelling. The proposed approach requires only an annotation-free peak table and produces an extremely reduced set of the most relevant features together with validation via Receiver Operating Characteristic analysis for selected predictors, cross-validation and unsupervised projection. The presented study was initially optimised by its own experimental set and then was successfully tested by using 36 datasets from 21 publicly available metabolomics projects. The suggested workflow can be used for classification purposes in high dimensional metabolomics studies and as a first step in exploratory analysis, data projection, biomarker selection, data integration and fusion.

MVCLN: Multi-View Convolutional LSTM Network for Cross-Media 3D Shape Recognition

Cross-media 3D model recognition is an important and challenging task in computer vision, which can be utilized in many applications such as landmark detection, image set classification, etc. In recent years, with the development of deep learning, many approaches have been proposed to handle the 3D model recognition problem. However, all of these methods focus on the structure information representation and the multi-view information fusion, and ignore the spatial and temporal information. So that it is not suitable for the cross-media 3D model recognition. In this paper, we utilize the sequence views to represent each 3D model and propose a novel Multi-view Convolutional LSTM Network (MVCLN), which utilizes the LSTM structure to extract temporal information and applies the convolutional operation to extract spatial information. More especially, the spatial and temporal information both are considered during the training process, which can effectively utilize the differences between the view's spatial information to improve the final performance. Meanwhile, we also introduce the classic attention model to define the weight of each view, which can reduce the redundant information of view's spatial information in the information fusion step. We evaluate the proposed method on the ModelNet40 for 3D model classification and retrieval task. We also construct a dataset utilizing the overlap categories of MV-RED, ShapenetCore and ModelNet to demonstrate the effectiveness of our approach for the cross-media 3D model recognition. Experimental results and comparisons with the state-of-the-art methods demonstrate that our framework can achieve superior performance.

Doxorubicin-caused cardiomyopathy is due to an inhibition of the fusion step on the autophagy flux

Doxorubicin-caused cardiomyopathy is due to an inhibition of the fusion step on the autophagy flux

Evaluating MERS-CoV Entry Pathways.

Middle East respiratory syndrome coronavirus (MERS-CoV) is an emerging zoonotic pathogen with a broad host range. The extent of MERS-CoV in nature can be traced to its adaptable cell entry steps. The virus can bind host-cell carbohydrates as well as proteinaceous receptors. Following receptor interaction, the virus can utilize diverse host proteases for cleavage activation of virus-host cell membrane fusion and subsequent genome delivery. The fusion and genome delivery steps can be completed at variable times and places, either at or near cell surfaces or deep within endosomes. Investigators focusing on the CoVs have developed several methodologies that effectively distinguish these different cell entry pathways. Here we describe these methods, highlighting virus-cell entry factors, entry inhibitors, and viral determinants that specify the cell entry routes. While the specific methods described herein were utilized to reveal MERS-CoV entry pathways, they are equally suited for other CoVs, as well as other protease-dependent viral species.

Multilayer feature selection method for polyp classification via computed tomographic colonography.

Polyp classification is a feature selection and clustering process. Picking the most effective features from multiple polyp descriptors without redundant information is a great challenge in this procedure. We propose a multilayer feature selection method to construct an optimized descriptor for polyp classification with a feature-grouping strategy in a hierarchical framework. First, the proposed method makes good use of image metrics, such as intensity, gradient, and curvature, to divide their corresponding polyp descriptors into several feature groups, which are the preliminary units of this method. Then each preliminary unit generates two ranked descriptors, i.e., their optimized variable groups (OVGs) and preliminary classification measurements. Next, a feature dividing-merging (FDM) algorithm is designed to perform feature merging operation hierarchically and iteratively. Unlike traditional feature selection methods, the proposed FDM algorithm includes two steps for feature dividing and feature merging. At each layer, feature dividing selects the OVG with the highest area under the receiver operating characteristic curve (AUC) as the baseline while other descriptors are treated as its complements. In the fusion step, the FDM merges some variables with gains into the baseline from the complementary descriptors iteratively on every layer until the final descriptor is obtained. This proposed model (including the forward step algorithm and the FDM algorithm) is a greedy method that guarantees clustering monotonicity of all OVGs from the bottom to the top layer. In our experiments, all the selected results from each layer are reported by both graphical illustration and data analysis. Performance of the proposed method is compared to five existing classification methods by a polyp database of 63 samples with pathological reports. The experimental results show that our proposed method outperforms other methods by 4% to 23% gains in terms of AUC scores.

Abstract B06: The angiotensin receptor blocker and partial PPARγ agonist telmisartan inhibits the growth of pancreatic ductal adenocarcinoma

Abstract Pancreatic ductal adenocarcinoma (PDAC) is highly resistant to chemotherapy, partly due to the presence of a dense-fibrotic stroma and adaptive metabolism. Telmisartan is an angiotensin II type receptor 1 (AT1) antagonist with partial peroxisome proliferator-activated receptor gamma (PPARγ) agonistic activity used for treatment of hypertension. The aim of this study was to determine the effects of telmisartan on the viability of PDAC cells and tumor progression. In panels of 4 murine and 8 human PDAC cells, the telmisartan IC50 was lower in cells with a low steady-state expression of PPARγ and a mesenchymal cell morphology. In contrast, losartan—a selective AT1 inhibitor—did not affect the viability of PDAC cells. The siRNA knockdown of PPARγ enhanced the sensitivity of telmisartan and stimulated epithelial-mesenchymal transition, which was accompanied by an increase in Wnt signaling. PPARγ regulates glucose metabolism and autophagy. We thus assessed effects of telmisartan on bioenergetics and autophagy of PDAC cells. In PPARγ-knockdown and -scrambled cells telmisartan significantly reduced glucose uptake, without affecting ATP production, but increased respiratory capacity, which can maintain the production of ATP during hypoglycemia. Immunoblotting revealed that PPARγ knockdown compared to scramble cells had increased levels of phosphorylated-AMP-activated protein kinase (p-AMPK) and increased expression of LC3A/B—structural proteins of autophagosomal membranes—which implies higher levels of autophagy. We also compared effects of telmisartan treatment on LC3A/B expression to well-established autophagy modulators, chloroquine and verapamil. Under nutrient-rich conditions and as expected, chloroquine and verapamil treatment induced LC3A/B accumulation, consistent with active autophagic flux in these cells. Telmisartan treatment decreased the levels of LC3A/B in both scramble and PPARγ knockdown cells and decreased the formation of LC3A/B positive granules in other PDAC cell lines. Telmisartan can also induce the accumulation of the signal adaptor protein p62 (SQSTM1), even in the presence of verapamil, which is also consistent with autophagy inhibition. Telmisartan did not prevent the accumulation LC3A/B in the presence of chloroquine, implying that telmisartan acts after the autophagosome-lysosome fusion step. To assess the effects of telmisartan in vivo, we used an orthotopic PDAC model. Telmisartan monotherapy inhibited the growth of primary tumors, decreased the incidence of liver metastasis, and significantly improved the survival of mice. Hence, telmisartan can reduce autophagy and the viability of PDAC cells, and PDAC progression. Because telmisartan is an FDA-approved drug, our findings provide the scientific rationale for testing its efficacy in the prevention of PDAC progression. Citation Format: Jelena Grahovac, Shiwei Han, Hao Liu, Mark Duquette, Alba Luengo, Daniel Schanne, Andrew S. Liss, Matthew G. Vander Heiden, Rakesh K. Jain, Yves Boucher. The angiotensin receptor blocker and partial PPARγ agonist telmisartan inhibits the growth of pancreatic ductal adenocarcinoma [abstract]. In: Proceedings of the AACR Special Conference on Pancreatic Cancer: Advances in Science and Clinical Care; 2019 Sept 6-9; Boston, MA. Philadelphia (PA): AACR; Cancer Res 2019;79(24 Suppl):Abstract nr B06.

Synthesis of zeolite A using the waste of iron mine tailings dam and its application for industrial effluent treatment

Abstract In November 2015, a dam from iron mining collapsed, which is considered as the largest environmental accident in Brazil. Over than 50 million m³ were released reaching the Doce River and killed 20 people in Mariana. The literature review shows that the mining tailings can be used as raw material to the synthesis of zeolites, which can be used for the wastewater treatment. The aim of this work was the synthesis of zeolites from the Samarco tailings and its application for the treatment of electroplating effluent. The residue was characterized which identified high-iron content for the synthesis of zeolites. The residue was mixed with NaOH (1:1) and the reaction temperature was evaluated from 350 °C to 650 °C. Then, the material was mixed with aluminum source and the effect of time was evaluated. The zeolite was applied to the wastewater treatment from electroplating industry. Results showed that the synthesis of zeolite A was carried out with a fusion step at 450 °C for 1 h using RAS: NaOH ratio 1:1, and further hydrothermal treatment at 100 °C during 4 h. The zinc removal from the wastewater was up to 98% using 50 mL of the solution, 2.5 g zeolite A, 60min and pH 6.4. The main metals presented in the solution were adsorbed up to 90% by the zeolite synthesized. Zinc adsorption by the zeolite fitted better on Langmuir isotherm. The zeolite was used four times in a row and zinc removal declined 98–68%.

Content based image retrieval using fusion of multilevel bag of visual words

Content based image retrieval (CBIR) is the art of finding visually and conceptually similar pictures to the given query picture. Usually, there is a semantic gap between low-level image features and high-level concepts perceived by viewers. Although features such as intensity and color enforce a good distinction between the images in terms of greater detail, they convey little semantic information. Therefore, employing higher-level features such as properties of regions and objects within the image could improve the retrieval performance. In this study, features are extracted at the pixel, region, object, and concept levels. The fusion step concatenates the four feature vectors and maps it to a lower-dimensional space using auto-encoders. The experiments confirm the efficiency of the proposed method over the individual feature groups and also the state of the art methods.

The Step of Incorporation of Bacillus coagulans GBI-30 6086 Into "requeijão cremoso" Processed Cheese Does Not Affect Metabolic Homeostasis of Rats.

Dairy product consumption is a common habit in Brazil. These products present a good matrix for probiotic incorporation. Thus, in this study the feasibility of producing a probiotic “requeijão cremoso” incorporated with Bacillus coagulans GBI-30 6086 in three different steps and its metabolic effect in an animal model for 2 weeks has been evaluated. Wistar adult health rats were randomized into one to five groups (n = 8 for each group): Control (C); “requeijão cremoso” without probiotic (RC); probiotic inoculated in the milk before pasteurization at 65°C/30 min (RPP); “requeijão cremoso” inoculated before the fusion step and consequently exposed to 90°C/5 min (RPF); and “requeijão cremoso” inoculated after fusion step, i.e., once the product temperature reached 50°C (RPAF). At the end of treatment, analysis of molecular markers of proteins of stress and antioxidant system, HSP 25, 60, 70 and 90, SOD and catalase were performed in the animals’ muscles by Western Blot technique. The HSP25, HSP90 and catalase levels of C, RPP, RPF, and RPAF were similar, indicating that the homeostasis remained unchanged. The incorporation of B. coagulans GBI-30 6086 in the “requeijão cremoso” was shown to be stable and the microorganism remained viable in all steps tested. The incorporation of the probiotic strain in the fusion stage facilitated the technological process, since it allowed a better homogenization of the product and did not affect the maintenance of the metabolic homeostasis of rats.

U18666A inhibits classical swine fever virus replication through interference with intracellular cholesterol trafficking

The level of cholesterol in host cells has been demonstrated to affect viral infection. Our previous studies showed that cholesterol-rich membrane rafts mediated the entry of classical swine fever virus (CSFV) into PK-15 or 3D4/21 cells, but the role of cholesterol post entry was still not clear. In this study, we found that CSFV replication before fusion was affected when the cholesterol trafficking in infected cells was disrupted using a cholesterol transport inhibitor, U18666A. Our data showed that U18666A affected both the fusion and replication steps in the life cycle of the virus, but not its binding and entry steps. The subsequent experiments confirmed that niemann-pick C1 (NPC1), a lysosomal membrane protein that helps cholesterol to leave the lysosome, was affected by U18666A, which led to the accumulation of cholesterol in lysosomes and inhibition of CSFV replication. Imipramine, a cationic hydrophobic amine similar to U18666A, also inhibited CSFV replication via similar mechanism. Surprisingly, the antiviral effect of U18666A was restored by the histone deacetylase inhibitor (HDACi), Vorinostat, which suggested that HDACi reverted the dysfunction of NPC1, and intra-cellular cholesterol accumulation disappeared and CSFV replicability resumed. Together, these data indicated that CSFV transformed from early endosome and late endosome into lysosome after endocytosis for further replication and that U18666A was a potential drug candidate for anti-pestivirus treatment.