Protein Engineering with Lightweight Graph Denoising Neural Networks.

Protein engineering faces challenges in finding optimal mutants from a massive pool of candidate mutants. In this study, we introduce a deep-learning-based data-efficient fitness prediction tool to steer protein engineering. Our methodology establishes a lightweight graph neural network scheme for protein structures, which efficiently analyzes the microenvironment of amino acids in wild-type proteins and reconstructs the distribution of the amino acid sequences that are more likely to pass natural selection. This distribution serves as a general guidance for scoring proteins toward arbitrary properties on any order of mutations. Our proposed solution undergoes extensive wet-lab experimental validation spanning diverse physicochemical properties of various proteins, including fluorescence intensity, antigen-antibody affinity, thermostability, and DNA cleavage activity. More than 40% of ProtLGN-designed single-site mutants outperform their wild-type counterparts across all studied proteins and targeted properties. More importantly, our model can bypass the negative epistatic effect to combine single mutation sites and form deep mutants with up to seven mutation sites in a single round, whose physicochemical properties are significantly improved. This observation provides compelling evidence of the structure-based model's potential to guide deep mutations in protein engineering. Overall, our approach emerges as a versatile tool for protein engineering, benefiting both the computational and bioengineering communities.

Construction and validation of artificial intelligence pathomics models for predicting pathological staging in colorectal cancer: Using multimodal data and clinical variables.

OBJECTIVE: This retrospective observational study aims to develop and validate artificial intelligence (AI) pathomics models based on pathological Hematoxylin-Eosin (HE) slides and pathological immunohistochemistry (Ki67) slides for predicting the pathological staging of colorectal cancer. The goal is to enable AI-assisted accurate pathological staging, supporting healthcare professionals in making efficient and precise staging assessments. METHODS: This study included a total of 267 colorectal cancer patients (training cohort: n = 213; testing cohort: n = 54). Logistic regression algorithms were used to construct the models. The HE image features were used to build the HE model, the Ki67 image features were used for the Ki67 model, and the combined model included features from both the HE and Ki67 images, as well as tumor markers (CEA, CA724, CA125, and CA242). The predictive results of the HE model, Ki67 model, and tumor markers were visualized through a nomogram. The models were evaluated using ROC curve analysis, and their clinical value was estimated using decision curve analysis (DCA). RESULTS: A total of 260 deep learning features were extracted from HE or Ki67 images. The AUC for the HE model and Ki67 model in the training cohort was 0.885 and 0.890, and in the testing cohort, it was 0.703 and 0.767, respectively. The combined model and nomogram in the training cohort had AUC values of 0.907 and 0.926, and in the testing cohort, they had AUC values of 0.814 and 0.817. In clinical DCA, the net benefit of the Ki67 model was superior to the HE model. The combined model and nomogram showed significantly higher net benefits compared to the individual HE model or Ki67 model. CONCLUSION: The combined model and nomogram, which integrate pathomics multi-modal data and clinical-pathological variables, demonstrated superior performance in distinguishing between Stage I-II and Stage III colorectal cancer. This provides valuable support for clinical decision-making and may improve treatment strategies and patient prognosis. Furthermore, the use of immunohistochemistry (Ki67) slides for pathomics modeling outperformed HE slide, offering new insights for future pathomics research.

Development and validation of a Radiopathomics model based on CT scans and whole slide images for discriminating between Stage I-II and Stage III gastric cancer.

OBJECTIVE: This study aimed to develop and validate an artificial intelligence radiopathological model using preoperative CT scans and postoperative hematoxylin and eosin (HE) stained slides to predict the pathological staging of gastric cancer (stage I-II and stage III). METHODS: This study included a total of 202 gastric cancer patients with confirmed pathological staging (training cohort: n = 141; validation cohort: n = 61). Pathological histological features were extracted from HE slides, and pathological models were constructed using logistic regression (LR), support vector machine (SVM), and NaiveBayes. The optimal pathological model was selected through receiver operating characteristic (ROC) curve analysis. Machine learnin algorithms were employed to construct radiomic models and radiopathological models using the optimal pathological model. Model performance was evaluated using ROC curve analysis, and clinical utility was estimated using decision curve analysis (DCA). RESULTS: A total of 311 pathological histological features were extracted from the HE images, including 101 Term Frequency-Inverse Document Frequency (TF-IDF) features and 210 deep learning features. A pathological model was constructed using 19 selected pathological features through dimension reduction, with the SVM model demonstrating superior predictive performance (AUC, training cohort: 0.949; validation cohort: 0.777). Radiomic features were constructed using 6 selected features from 1834 radiomic features extracted from CT scans via SVM machine algorithm. Simultaneously, a radiopathomics model was built using 17 non-zero coefficient features obtained through dimension reduction from a total of 2145 features (combining both radiomics and pathomics features). The best discriminative ability was observed in the SVM_radiopathomics model (AUC, training cohort: 0.953; validation cohort: 0.851), and clinical decision curve analysis (DCA) demonstrated excellent clinical utility. CONCLUSION: The radiopathomics model, combining pathological and radiomic features, exhibited superior performance in distinguishing between stage I-II and stage III gastric cancer. This study is based on the prediction of pathological staging using pathological tissue slides from surgical specimens after gastric cancer curative surgery and preoperative CT images, highlighting the feasibility of conducting research on pathological staging using pathological slides and CT images.

Heterogeneous integration of an on-chip Nd:YAG whispering gallery mode laser with a lithium-niobate-on-insulator platform.

The integration of heterogeneous optical components onto an optical platform is crucial for the advancement of photonic chips. To achieve this, efficient coupling of optical signals between components and the platform is essential. Here, we have successfully integrated a Nd:YAG microdisk laser with a lithium-niobate-on-insulator (LNOI) photonic platform by modulating the propagation modes of LNOI. Ridge waveguides are fabricated on the LNOI by carefully adjusting the cross-sectional dimensions to enable the propagation of higher-order propagation modes. This ridge waveguide ensures that the effective refractive index of the higher-order mode closely matches that of the fundamental mode of the Nd:YAG microdisk, ensuring efficient waveguide-microdisk coupling. This on-chip laser, consisting of an Nd:YAG microdisk and LNOI integration, achieves a maximum output power of 23 µW, and a mode suppression ratio of 53.6 dB. This research presents an efficient approach for constructing highly functional heterogeneous integrated optical chips.

Low energy consumption layout of exhibition buildings in Yangtze River Delta region.

The issue of high energy consumption in exhibition buildings has long been a focal point in the field of architectural design. However, current energy consumption assessments for exhibition buildings mainly focus on post-construction evaluations, lacking corresponding guidance during the initial design phase. To address this issue, this study selected 48 well-known exhibition buildings both domestically and internationally as research subjects. Utilizing scatter plot curve fitting, it was discovered that there exists a nonlinear quadratic relationship between the building area of the first floor and the courtyard area. Based on this relationship, four typical layouts were established to match the climatic characteristics of Hangzhou, a representative region in the Yangtze River Delta of China. Taking into account regional architectural features, the study specifically examined the impact of different orientations and window-to-wall ratios on energy consumption levels. The influence of these factors on energy consumption was analyzed using the DesignBuilder software. The results revealed that there exists an optimal window-to-wall ratio for exhibition buildings, with parallel, L-shaped enclosed south-facing courtyards, and U-shaped enclosed east-facing courtyards showing greater energy efficiency. This research provides guidance for designing exhibition buildings that are energy-efficient and foster a harmonious indoor-outdoor relationship.

Selected legacy and emerging organic contaminants in sediments of China's Yangtze - the world's third longest river: Response to anthropogenic activities.

To explore contaminant concerns as a result of anthropogenic disturbance of the river system, this study provided the first extensive investigation of the contamination profiles, possible driving factors, and ecological risks of 40 target compounds including pharmaceuticals and personal care products (PPCPs), neonicotinoid pesticides (NNIs), polybrominated diphenyl ethers (PBDEs), and polychlorinated biphenyls (PCBs) in sediments of the whole Yangtze River (the world's third longest river). Among these target compounds, PPCPs were the dominant contaminants with a total concentration (∑15PPCPs) of 2.13-14.99 ng/g, followed by ∑7PCBs (

Analysis of the potential application of a residential composite energy storage system based on a double-layer optimization model.

Along with the further integration of demand management and renewable energy technology, making optimal use of energy storage devices and coordinating operation with other devices are key. The present study takes into account the current situation of power storage equipment. Based on one year of measured data, four cases are designed for a composite energy storage system (ESS). In this paper, a two-tiered optimization model is proposed and is used to optimizing the capacity of power storage devices and the yearly production of the system. Furthermore, this paper performs a comparative analysis of the performance of the four cases from the energy, environmental and economic perspectives. It is concluded that this kind of energy storage equipment can enhance the economics and environment of residential energy systems. The thermal energy storage system (TESS) has the shortest payback period (7.84 years), and the CO2 emissions are the lowest. Coupled with future price volatility and the carbon tax, the electrothermal hybrid energy storage system (HESS) has good development potential. However, the current investment cost is very high, and it will not be possible to recover this cost in 10 years. Finally, it is recommended that the cost of equipment be reduced in combination with subsidies and incentives for further promotion. The research results not only fill a gap in the study area, but also provide some suggestions for further development of industry and research on user-side energy storage.

MedChatZH: A tuning LLM for traditional Chinese medicine consultations.

Generative Large Language Models (LLMs) have achieved significant success in various natural language processing tasks, including Question-Answering (QA) and dialogue systems. However, most models are trained on English data and lack strong generalization in providing answers in Chinese. This limitation is especially evident in specialized domains like traditional Chinese medical QA, where performance suffers due to the absence of fine-tuning and high-quality datasets. To address this, we introduce MedChatZH, a dialogue model optimized for Chinese medical QA based on transformer decoder with LLaMA architecture. Continued pre-training on a curated corpus of Chinese medical books is followed by fine-tuning with a carefully selected medical instruction dataset, resulting in MedChatZH outperforming several Chinese dialogue baselines on a real-world medical dialogue dataset. Our model, code, and dataset are publicly available on GitHub (https://github.com/tyang816/MedChatZH) to encourage further research in traditional Chinese medicine and LLMs.

Remazolam affects the phenotype and function of astrocytes to improve traumatic brain injury by regulating the Cx43.

PURPOSE: To explore the mechanism by which Remazolam affects the phenotype and function of astrocytes to improve traumatic brain injury (TBI). METHODS: The oxygen -glucose deprivation/recovery (OGD/R) cell model was constructed to simulate the pathological state of astrocytes in a TBI environment. The viability of astrocytes was measured by CCK-8, and the cytoskeleton changes were observed by Phalloidin- TRITC staining. The expressions of differentiation markers, Cx43 and phosphorylated Cx43 (P-Cx43) of A1/A2 astrocytes were detected by Western blot, and the complement C3 and S100A10 of A1/A2 astrocytes were detected by ELISA. The TBI rat model was established. The water content of brain tissue was measured by dry-wet specific gravity method, the pathological morphology of brain tissue in cortical injury area was observed by HE staining method, ROS was detected by fluorescence quantitative method, Cx43 expression was detected by immunohistochemistry method, and the differentiation markers of A1/A2 astrocytes were detected by immunofluorescence. RESULTS: In the TBI environment, astrocytes showed decreased cell viability, blurred skeleton, and increased expression of Cx43. In TBI rats, the water content of brain tissue increased, the brain tissue in the cortex injury area was seriously damaged, ROS and Cx43 expression were significantly increased, and mainly distributed in A2 astrocytes. Remazolam can reverse the above results after the intervention. CONCLUSION: Remazolam affects the phenotype and function of astrocytes to improve TBI via regulating Cx43, and plays a role in protecting the neurological function of TBI rats.

Occurrence, source apportionment and risk assessment of perfluorinated compounds in sediments from the longest river in Asia.

This study investigated the pollution of perfluorinated compounds (PFCs) in sediments from the main stream of the Yangtze River, the longest river in Asia. Totally, 13 of 15 PFASs were detected in the sediments and the total concentrations ranged from 0.058 ng/g to 0.89 ng/g dry weight (dw), with dominant contaminants by perfluorooctane sulfonic acid (PFOS) and perfluorooctanoic acid (PFOA). Concentrations of PFASs in the downstream were higher than those of upstream and midstream. Four main sources were analysed using the Unmix model, textile treatments and food packaging dominantly accounted for approximately half of the total sources, followed by metal electroplating (26.8%), fluoropolymer products (16.3%) and fluororesin coatings (7.4%). Total organic carbon (TOC), total nitrogen (TN) and grain size had significant correlation with the concentration of PFASs in sediments, indicating that the physical and chemical parameters could directly affect the adsorption process of PFASs. In addition, anthropogenic factors such as urbanization rate and per capita GDP also had a direct impact on the distribution of PFASs. Environmental risk assessment showed that PFOS posed low to medium risks to the Yangtze River, indicating that sustained attentions were needed.

Ultrasound Microbubble-Stimulated miR-145-5p Inhibits Malignant Behaviors of Breast Cancer Cells by Targeting ACTG1.

ABSTRACT: Ultrasound-targeted microbubble destruction (UTMD) technology combines ultrasound with a variety of functional microbubble vectors to enhance the transfection and expression of target genes, and has become a promising noninvasive method for localized gene transfer, which is widely used in gene therapy for cancer. This research aimed to explore the role of UTMD-mediated miR-145-5p on breast cancer (BC) tumorigenesis and the underlying mechanisms. To achieve UTMD-mediated miR-145-5p overexpression, BC cells were cotransfected with microbubbles (MBs) and miR-145-5p mimics. The BC cell malignant phenotypes were assessed through CCK-8, wound healing, and transwell assays. MiR-145-5p and actin gamma 1 (ACTG1) binding relationship was verified through luciferase reporter and RNA pull-down assays. MiR-145-5p and ACTG1 levels in BC cells and tissues were detected through RT-qPCR and Western blotting. ACTG1 was upregulated, whereas miR-145-5p was downregulated in BC cells and tissues. MiR-145-5p targeted ACTG1 and negatively regulated its level in BC cells. Overexpressing miR-145-5p restrained BC cell growth, migration, and invasion. Ultrasound-targeted microbubble destruction improved the overexpression efficiency of miR-145-5p and enhanced the suppressive influence on BC cell malignant phenotypes. In addition, ACTG1 overexpression compromises the repression of UTMD-mediated miR-145-5p on cellular behaviors in BC. Ultrasound-targeted microbubble destruction-delivered miR-145-5p hindered malignant behaviors of BC cells through downregulating ACTG1.

Biomimetic construction of environmental-tolerant composite hydrogels based on galactomannan for tough, flexible and conductive sensors.

Sustainable composite hydrogel materials with harsh environmental adaption and tolerance capability have received considerable interests but still remain as challenges. In this work, biomimetic strategy was adapted for construction of three-dimensional galactomannan (GM) hydrogels with intercalation of flexible polymer chains polyethyleneimine (PEI), biomacromolecules tannin acid (TA) and CeO2 nanoparticles (NPs). The hydrogels cross-linked with double-networks (DN) present not only pH-responsive water absorption property, but also boosted mechanical strength with highest toughness of 326 kJ/m3 and Young's modulus of 220 kPa. Self-healing and anti-freezing capabilities were revealed for the hydrogels by maintaining of fracture elongation (23 %) and fracture strength (250 kPa). TA, CeO2 NPs as well as the amide groups in PEI of the hydrogels introduced excellent bacterial prohibition performance on both Bacillus subtilis (B. subtilis) and Escherichia coli (E. coli). Also, due to the existence of the free ions, the hydrogels exhibited electric conductive properties, with wide-range high sensitivity and long-time conductive stability. In addition, various tensile strain degrees were related to the conductive resistance values, and the great recovery performance was proved by cyclic tensile-conductive tests for 3000 times. Therefore, the proposed GM-based hydrogels displayed great potentials as strain sensors that are adaptable and tolerant to various environmental conditions.

Transferability-Guided Cross-Domain Cross-Task Transfer Learning.

We propose two novel transferability metrics fast optimal transport-based conditional entropy (F-OTCE) and joint correspondence OTCE (JC-OTCE) to evaluate how much the source model (task) can benefit the learning of the target task and to learn more generalizable representations for cross-domain cross-task transfer learning. Unlike the original OTCE metric that requires evaluating the empirical transferability on auxiliary tasks, our metrics are auxiliary-free such that they can be computed much more efficiently. Specifically, F-OTCE estimates transferability by first solving an optimal transport (OT) problem between source and target distributions and then uses the optimal coupling to compute the negative conditional entropy (NCE) between the source and target labels. It can also serve as an objective function to enhance downstream transfer learning tasks including model finetuning and domain generalization (DG). Meanwhile, JC-OTCE improves the transferability accuracy of F-OTCE by including label distances in the OT problem, though it incurs additional computation costs. Extensive experiments demonstrate that F-OTCE and JC-OTCE outperform state-of-the-art auxiliary-free metrics by 21.1% and 25.8% , respectively, in correlation coefficient with the ground-truth transfer accuracy. By eliminating the training cost of auxiliary tasks, the two metrics reduce the total computation time of the previous method from 43 min to 9.32 and 10.78 s, respectively, for a pair of tasks. When applied in the model finetuning and DG tasks, F-OTCE shows significant improvements in the transfer accuracy in few-shot classification experiments, with up to 4.41% and 2.34% accuracy gains, respectively.

Photosensitizer-loaded hydrogels: A new antibacterial dressing.

Bacterial wound infection has emerged as a pivotal threat to human health worldwide, and the situation has worsened owing to the gradual increase in antibiotic-resistant bacteria caused by the improper use of antibiotics. To reduce the use of antibiotics and avoid the increase in antibiotic-resistant bacteria, researchers are increasingly paying attention to  photodynamic therapy, which uses light to produce reactive oxygen species to kill bacteria. Treating bacteria-infected wounds by photodynamic therapy requires fixing the photosensitizer (PS) at the wound site and maintaining a certain level of wound humidity. Hydrogels are materials with a high water content and are well suited for fixing PSs at wound sites for antibacterial photodynamic therapy. Therefore, hydrogels are often loaded with PSs for treating bacteria-infected wounds via antibacterial photodynamic therapy. In this review, we systematically summarised the antibacterial mechanisms and applications of PS-loaded hydrogels for treating bacteria-infected wounds via photodynamic therapy. In addition, the recent  studies and the research status progresses of novel antibacterial hydrogels are discussed. Finally, the challenges and future prospects of PS-loaded hydrogels are reviewed.

Different clinical diagnosis end up in the same pathological diagnosis of intravascular leiomyomatosis: Two case reports.

RATIONALE: Intravascular/intravenous leiomyomatosis (IVL) is a peculiar variant of uterine leiomyoma that is classified as a histologically benign smooth muscle tumor with a biological behavior similar to that of a malignant tumor. It is characterized by the proliferation of leiomyomas spreading along the uterine and extrauterine venous circulation. PATIENT CONCERNS: Herein, we present 2 cases of IVL who had completely different clinical manifestations to stress the need for constant vigilance of IVL diagnosis and the understanding of uterine leiomyoma heterogenicity. Case 1 was registered for fever without specific triggering factors, irregular menstruation and clinically diagnosed uterine diverticula, while no information about fibroids was mentioned. Case 2 was characterized by an aggressively growing abdominal mass. With a large space-occupying lesion in the right abdominopelvic cavity and no imaging evidence of involvement of the iliac vein or above vein, the patient was initially diagnosed with multiple myomata. DIAGNOSES: Both patients' diagnoses were confirmed as IVL by histopathology. To our knowledge, the mass of case 1 is the minimum IVL in the English literature. INTERVENTIONS: Subtotal hysterectomy with bilateral salpingectomy was performed on the former, while total hysterectomy with bilateral salpingectomy was performed on the latter. OUTCOMES: Both patients were comfortable, and no relapse occurred. LESSONS: Two cases in the study showed 2 different proceeding stages of the same disease and corroborated multiple pathogeneses, which have been mentioned in the available literature on IVL. Our work provides both supplement for clinical data to facilitate further research and better understanding of special types of fibroids to clinicians.

Dose- and stage-dependent toxic effects of prenatal prednisone exposure on fetal articular cartilage development.

Prednisone is frequently used to treat rheumatoid diseases in pregnant women because of its high degree of safety. Whether prenatal prednisone exposure (PPE) negatively impacts fetal articular cartilage development is unclear. In this study, we simulated a clinical prednisone treatment regimen to examine the effects of different timings and doses of PPE on cartilage development in female and male fetal mice. Prednisone doses (0.25, 0.5, and 1 mg/kg/d) was administered to Kunming mice at different gestational stages (0-9 gestational days, GD0-9), mid-late gestation (GD10-18), or during the entire gestation (GD0-18) by oral gavage. The amount of matrix aggrecan (ACAN) and collagen type II a1(COL2a1), and expression of transforming growth factor ß1 (TGFß1) signaling pathway also demonstrated that the chondrocyte count and ACAN and COL2a1 expression reduced in fetal mice with early and mid-late PPE, with the reduction being more significant in the mice with early PPE than that in those with PPE at other stages. Prenatal exposure to different prednisone doses prevented the reduction of TGFß signaling pathway-related genes [TGFßR1, SMAD family member 3 (Smad3), SRY-box9 (SOX9)] as well as ACAN and COL2a1 mRNA expression levels in fetal mouse cartilage, with the most significant decrease after 1 mg/kg·d PPE. In conclusion, PPE can inhibit/restrain fetal cartilage development, with the greatest effect at higher clinical dose (1 mg/kg·d) and early stage of pregnancy (GD0-9), and the mechanism may be related to TGFß signaling pathway inhibition. The result of this study provide a theoretical and experimental foundation for the rational clinical use of prednisone.

Salvianolic Acid B Alleviates Liver Injury by Regulating Lactate-Mediated Histone Lactylation in Macrophages.

Salvianolic acid B (Sal B) is the primary water-soluble bioactive constituent derived from the roots of Salvia miltiorrhiza Bunge. This research was designed to reveal the potential mechanism of Sal B anti-liver injury from the perspective of macrophages. In our lipopolysaccharide-induced M1 macrophage model, Sal B showed a clear dose-dependent gradient of inhibition of the macrophage trend of the M1 type. Moreover, Sal B downregulated the expression of lactate dehydrogenase A (LDHA), while the overexpression of LDHA impaired Sal B's effect of inhibiting the trend of macrophage M1 polarization. Additionally, this study revealed that Sal B exhibited inhibitory effects on the lactylation process of histone H3 lysine 18 (H3K18la). In a ChIP-qPCR analysis, Sal B was observed to drive a reduction in H3K18la levels in the promoter region of the LDHA, NLRP3, and IL-1ß genes. Furthermore, our in vivo experiments showed that Sal B has a good effect on alleviating CCl4-induced liver injury. An examination of liver tissues and the Kupffer cells isolated from those tissues proved that Sal B affects the M1 polarization of macrophages and the level of histone lactylation. Together, our data reveal that Sal B has a potential mechanism of inhibiting the histone lactylation of macrophages by downregulating the level of LDHA in the treatment of liver injury.

Recent progress in engineering Clostridium autoethanogenum to synthesize the biochemicals and biocommodities.

Excessive mining and utilization fossil fuels has led to drastic environmental consequences, which will contribute to global warming and cause further climate change with severe consequences for the human population. The magnitude of these challenges requires several approaches to develop sustainable alternatives for chemicals and fuels production. In this context, biological processes, mainly microbial fermentation, have gained particular interest. For example, autotrophic gas-fermenting acetogenic bacteria are capable of converting CO, CO2 and H2 into biomass and multiple metabolites through Wood-Ljungdahl pathway, which can be exploited for large-scale fermentation processes to sustainably produce bulk biochemicals and biofuels (e.g. acetate and ethanol) from syngas. Clostridium autoethanogenum is one representative of these chemoautotrophic bacteria and considered as the model for the gas fermentation. Recently, the development of synthetic biology toolbox for this strain has enabled us to study and genetically improve their metabolic capability in gas fermentation. In this review, we will summarize the recent progress involved in the understanding of physiological mechanism and strain engineering for C. autoethanogenum, and provide our perspectives on the future development about the basic biology and engineering biology of this strain.

Successful treatment of refractory ascites in a patient with liver cirrhosis combined with hepatic artery-portal vein malformation: A case report.

INTRODUCTION: Hepatic artery-portal vein malformation is rarely encountered in clinical practice. Here, we reported a case of liver cirrhosis combined with hepatic artery-portal vein malformation with refractory ascites as the main symptom. And it was successfully treated by us. The present case demonstrates the role of hepatic artery-portal vein malformation in cirrhotic ascites and the importance of early diagnosis and interventional treatment. This article may provides some experience for the treatment of such patients. PATIENT CONCERNS: The patient was a 72-year-old woman with a 40-year history of Hepatitis B virus surface antigen positivity who sought medical advice with a chief complaint of abdominal distension for 1 week. DIAGNOSES: Enhanced abdominal computed tomography imaging of this patient revealed liver cirrhosis, splenomegaly, esophageal and gastric varices, massive ascites, and a low-density area in the S4 segment of the liver with an ambiguous boundary. Widening of the left branch of the portal vein was evident, and the portal vein was highlighted in the arterial phase and the venous phase. Digital subtraction angiography revealed substantial thickening of the left hepatic artery, and the administered contrast agent drained through the malformed vascular mass to the thickened left portal vein. Liver cirrhosis combined with hepatic artery-portal vein malformation were diagnosed. And we considered that the artery-portal vein malformation in this patient might be caused by cirrhosis. INTERVENTIONS: The patient was applied diuretics, entecavir and transcatheter embolization. OUTCOMES: The patient ascites did not resolve significantly when treated with diuretics alone. After the transcatheter embolization, the patient ascites relieved remarkably. CONCLUSION: The patient underwent transcatheter embolization for hepatic artery-portal vein malformation, after which her ascites resolved with good short-term curative efficacy. So, the patients who suffered from liver cirrhosis combined with hepatic artery-portal vein malformation and refractory ascites, should be active on transcatheter embolization.