Employing bimodal representations to predict DNA bendability within a self-supervised pre-trained framework.

The bendability of genomic DNA, which measures the DNA looping rate, is crucial for numerous biological processes of DNA. Recently, an advanced high-throughput technique known as 'loop-seq' has made it possible to measure the inherent cyclizability of DNA fragments. However, quantifying the bendability of large-scale DNA is costly, laborious, and time-consuming. To close the gap between rapidly evolving large language models and expanding genomic sequence information, and to elucidate the DNA bendability's impact on critical regulatory sequence motifs such as super-enhancers in the human genome, we introduce an innovative computational model, named MIXBend, to forecast the DNA bendability utilizing both nucleotide sequences and physicochemical properties. In MIXBend, a pre-trained language model DNABERT and convolutional neural network with attention mechanism are utilized to construct both sequence- and physicochemical-based extractors for the sophisticated refinement of DNA sequence representations. These bimodal DNA representations are then fed to a k-mer sequence-physicochemistry matching module to minimize the semantic gap between each modality. Lastly, a self-attention fusion layer is employed for the prediction of DNA bendability. In conclusion, the experimental results validate MIXBend's superior performance relative to other state-of-the-art methods. Additionally, MIXBend reveals both novel and known motifs from the yeast. Moreover, MIXBend discovers significant bendability fluctuations within super-enhancer regions and transcription factors binding sites in the human genome.

Prediction of biomarker-disease associations based on graph attention network and text representation.

MOTIVATION: The associations between biomarkers and human diseases play a key role in understanding complex pathology and developing targeted therapies. Wet lab experiments for biomarker discovery are costly, laborious and time-consuming. Computational prediction methods can be used to greatly expedite the identification of candidate biomarkers. RESULTS: Here, we present a novel computational model named GTGenie for predicting the biomarker-disease associations based on graph and text features. In GTGenie, a graph attention network is utilized to characterize diverse similarities of biomarkers and diseases from heterogeneous information resources. Meanwhile, a pretrained BERT-based model is applied to learn the text-based representation of biomarker-disease relation from biomedical literature. The captured graph and text features are then integrated in a bimodal fusion network to model the hybrid entity representation. Finally, inductive matrix completion is adopted to infer the missing entries for reconstructing relation matrix, with which the unknown biomarker-disease associations are predicted. Experimental results on HMDD, HMDAD and LncRNADisease data sets showed that GTGenie can obtain competitive prediction performance with other state-of-the-art methods. AVAILABILITY: The source code of GTGenie and the test data are available at: https://github.com/Wolverinerine/GTGenie.

MIX-TPI: a flexible prediction framework for TCR-pMHC interactions based on multimodal representations.

MOTIVATION: The interactions between T-cell receptors (TCR) and peptide-major histocompatibility complex (pMHC) are essential for the adaptive immune system. However, identifying these interactions can be challenging due to the limited availability of experimental data, sequence data heterogeneity, and high experimental validation costs. RESULTS: To address this issue, we develop a novel computational framework, named MIX-TPI, to predict TCR-pMHC interactions using amino acid sequences and physicochemical properties. Based on convolutional neural networks, MIX-TPI incorporates sequence-based and physicochemical-based extractors to refine the representations of TCR-pMHC interactions. Each modality is projected into modality-invariant and modality-specific representations to capture the uniformity and diversities between different features. A self-attention fusion layer is then adopted to form the classification module. Experimental results demonstrate the effectiveness of MIX-TPI in comparison with other state-of-the-art methods. MIX-TPI also shows good generalization capability on mutual exclusive evaluation datasets and a paired TCR dataset. AVAILABILITY AND IMPLEMENTATION: The source code of MIX-TPI and the test data are available at: https://github.com/Wolverinerine/MIX-TPI.

Hospital-treated infectious diseases, infection burden and risk of Parkinson disease: An observational and Mendelian randomization study.

BACKGROUND: Experimental and cross-sectional evidence has suggested a potential role of infection in the ethology of Parkinson's disease (PD). We aim to examine the longitudinal association of infections with the incidence of PD and to explore whether the increased risk is limited to specific infection type rather than infection burden. METHODS: Based on the UK Biobank, hospital-treated infectious diseases and incident PD were ascertained through record linkage to national hospital inpatient registers. Infection burden was defined as the sum of the number of infection episodes over time and the number of co-occurring infections. The polygenic risk score (PRS) for PD was calculated. The genome-wide association studies (GWAS) used in two-sample Mendelian Randomization (MR) were obtained from observational cohort participants of mostly European ancestry. RESULTS: Hospital-treated infectious diseases were associated with an increased risk of PD (adjusted HR [aHR] 1.35 [95 % CI 1.20-1.52]). This relationship persisted when analyzing new PD cases occurring more than 10 years post-infection (aHR 1.22 [95 % CI 1.04-1.43]). The greatest PD risk was observed in neurological/eye infection (aHR 1.72 [95 % CI 1.32-2.34]), with lower respiratory tract infection (aHR 1.43 [95 % CI 1.02-1.99]) ranked the second. A dose-response association was observed between infection burden and PD risk within each PD-PRS tertile (p-trend < 0.001). Multivariable MR showed that bacterial and viral infections increase the PD risk. CONCLUSIONS: Both observational and genetic analysis suggested a causal association between infections and the risk of developing PD. A dose-response relationship between infection burden and incident PD was revealed.

Magnetic field improves ozone production in an atmospheric pressure surface dielectric barrier discharge: understanding the physico-chemical mechanism behind low energy consumption.

Herein, we studied the combined effects of the magnetic field and the alternating current driven air surface dielectric barrier discharge on ozone production, and found that a 0.13 T perpendicular magnetic field introduced into the discharge area significantly enhanced the ozone generation performance with a 36-108% increase in ozone number density and 24-80% increase in ozone yield depending on discharge voltage and frequency differences. To reveal the micro physico-chemical mechanism of the influence of a magnetic field and discharge parameters of discharge voltage and frequency on ozone generation, a plasma chemical reaction network involving electron collision-chain reactions was considered. The results show that these parameters jointly influence ozone generation by affecting electron collision reactions and chain chemical reactions by changing the mean electron energy and plasma gas temperature. In this study, both the experimental results and mechanism analysis suggested that an optimal discharge parameter for ozone generation is the magnetic field assisted, low frequency, high voltage (6.5 kHz, 6.5 kV) surface dielectric barrier discharge. These insights provide guidance for optimizing the discharge parameters of the magnetic field assisted discharge to increase ozone production and reduce energy consumption.

Disrupted topological organization of functional brain networks is associated with cognitive impairment in hypertension patients: a resting-state fMRI study.

PURPOSE: To investigate the alterations of topological organization of the whole brain functional networks in hypertension patients with cognitive impairment (HTN-CI) and characterize its relationship with cognitive scores. METHODS: Fifty-seven hypertension patients with cognitive impairment and 59 hypertension patients with normal cognition (HTN-NC), and 49 healthy controls (HCs) underwent resting-state functional magnetic resonance imaging. Graph theoretical analysis was used to investigate the altered topological organization of the functional brain networks. The global topological properties and nodal metrics were compared among the three groups. Network-based statistic (NBS) analysis was used to determine the connected subnetwork. The relationships between network metrics and cognitive scores were also characterized. RESULTS: HTN-CI patients exhibited significantly decreased global efficiency, lambda, and increased shortest path length when compared with HCs. In addition, both HTN-CI and HTN-NC groups exhibited altered nodal degree centrality and nodal efficiency in the right precentral gyrus. The disruptions of global network metrics (lambda, Lp) and the nodal metrics (degree centrality and nodal efficiency) in the right precentral gyrus were positively correlated with the MoCA scores in HTN-CI. NBS analysis demonstrated that decreased subnetwork connectivity was present both in the HTN-CI and HTN-NC groups, which were mainly involved in the default mode network, frontoparietal network, and cingulo-opercular network. CONCLUSION: This study demonstrated the alterations of topographical organization and subnetwork connectivity of functional brain networks in HTN-CI. In addition, the global and nodal network properties were correlated with cognitive scores, which may provide useful insights for the understanding of neuropsychological mechanisms underlying HTN-CI.

Secure Grasping Detection of Objects in Stacked Scenes Based on Single-Frame RGB Images.

Secure grasping of objects in complex scenes is the foundation of many tasks. It is important for robots to autonomously determine the optimal grasp based on visual information, which requires reasoning about the stacking relationship of objects and detecting the grasp position. This paper proposes a multi-task secure grasping detection model, which consists of the grasping relationship network (GrRN) and the oriented rectangles detection network CSL-YOLO, which uses circular smooth label (CSL). GrRN uses DETR to solve set prediction problems in object detection, enabling end-to-end detection of grasping relationships. CSL-YOLO uses classification to predict the angle of oriented rectangles, and solves the angle distance problem caused by classification. Experiments on the Visual Manipulate Relationship Dataset (VMRD) and the grasping detection dataset Cornell demonstrate that our method outperforms existing methods and exhibits good applicability on robot platforms.

Reduced electric field and gas temperature effects on chemical product dynamics in air surface dielectric barrier discharges: from macro-physical parameters to micro-chemical mechanisms.

To gain insights into the mechanisms of plasma chemical product interactions, the dynamic changes of the surface dielectric barrier discharge (SDBD) products are experimentally related to the reduced electric field and gas temperature. The higher applied voltage and frequency cause faster product changes from the O3-containing to the O3-free state, while raising the electron energy and gas temperature. The electron energy affects the electron collision reactions and the production of various reactive species, steering the chemical reactions towards the predominant production of NO over O3. The gas temperature affects the generation and quenching rates of the key products. Collectively, this work bridges macro-physical parameters and micro-chemical mechanisms through the electron energy and gas temperature effects, and contributes to better understanding of the physico-chemical processes in low-temperature plasmas.

Prognostic factors in hepatocellular carcinoma patients with Child-Pugh A liver function after hepatectomy: Not related to the surgical approach.

INTRODUCTION AND OBJECTIVES: Improving the prognosis of patients with hepatocellular carcinoma (HCC) undergoing hepatectomy is critical. This article aims to investigate the risk factors affecting the prognosis of HCC patients with Child-Pugh A (CPA) liver function after hepatectomy and to compare the prognosis of patients with anatomical resection (AR) and nonanatomical resection (NAR). METHODS: In total, 186 patients diagnosed with HCC between 2013 and 2019 were retrospectively enrolled. Univariate and multivariate analyses were performed using a Cox proportional hazard regression model to explore the factors related to prognosis. Overall survival (OS) and progression-free survival (PFS) were analyzed by log-rank tests and are shown by Kaplan-Meier curves. Chi-square tests and Mann-Whitney U tests were used to compare the difference in clinical characteristics between AR and NAR patients. RESULTS: Among the 186 enrolled patients, only 73 were followed over 60 months. The 1-, 3-, and 5-year survival rates were 74.5%, 46.7% and 26.0%, respectively. Multivariate analyses demonstrated that portal vein invasion (PVI) and tumor size were independent risk factors for OS and PFS. Preoperative hepatitis B surface antigen (HBsAg) and a-fetoprotein (AFP) levels were identified as independent risk factors only for PFS. In univariate analysis, the NAR group had a better OS rate than the AR group (1-year: 80.4% vs. 63.6%, 3-year: 55.9% vs. 30.3%, 5-year: 34.8% vs. 11.1%), but this was not confirmed by multivariate analysis. CONCLUSIONS: PVI and tumor size > 5 cm are risk factors for the prognosis of CPA HCC patients after hepatectomy, but the surgical type is not.

BMRMI Reduces Depressive Rumination Possibly through Improving Abnormal FC of Dorsal ACC.

Rumination is a common symptom of major depressive disorder (MDD) and has been characterized as a vulnerability factor for the onset or recurrence of MDD. However, the neurobiological mechanisms underlying rumination and appropriate treatment strategies remain unclear. In the current study, we used resting-state functional magnetic resonance imaging to investigate the effects of body-mind relaxation meditation induction (BMRMI) intervention in MDD with rumination. To this aim, we have recruited 25 MDD and 24 healthy controls (HCs). Changes in functional connectivity (FC) of the anterior cingulate cortex (ACC) subregion and the scores of clinical measurements were examined using correlation analysis. At baseline, MDD showed stronger FC between the right dorsal ACC (dACC) and right superior frontal gyrus than did the HC group. Compared to baseline, the HC group showed a significantly enhanced FC between the right dACC and right superior frontal gyrus, and the MDD group demonstrated a significantly weaker FC between the left dACC and right middle frontal gyrus (MFG) after the intervention. Furthermore, the FC between the right dACC and right superior frontal gyrus was positively associated with rumination scores across all participants at baseline. The above results indicate that BMRMI may regulate self-referential processing and cognitive function through modulating FC of the dACC in MDD with rumination.

An Outline of Multi-Sensor Fusion Methods for Mobile Agents Indoor Navigation.

Indoor autonomous navigation refers to the perception and exploration abilities of mobile agents in unknown indoor environments with the help of various sensors. It is the basic and one of the most important functions of mobile agents. In spite of the high performance of the single-sensor navigation method, multi-sensor fusion methods still potentially improve the perception and navigation abilities of mobile agents. This work summarizes the multi-sensor fusion methods for mobile agents' navigation by: (1) analyzing and comparing the advantages and disadvantages of a single sensor in the task of navigation; (2) introducing the mainstream technologies of multi-sensor fusion methods, including various combinations of sensors and several widely recognized multi-modal sensor datasets. Finally, we discuss the possible technique trends of multi-sensor fusion methods, especially its technique challenges in practical navigation environments.

Real-Time 3D Reconstruction Method Based on Monocular Vision.

Real-time 3D reconstruction is one of the current popular research directions of computer vision, and it has become the core technology in the fields of virtual reality, industrialized automatic systems, and mobile robot path planning. Currently, there are three main problems in the real-time 3D reconstruction field. Firstly, it is expensive. It requires more varied sensors, so it is less convenient. Secondly, the reconstruction speed is slow, and the 3D model cannot be established accurately in real time. Thirdly, the reconstruction error is large, which cannot meet the requirements of scenes with accuracy. For this reason, we propose a real-time 3D reconstruction method based on monocular vision in this paper. Firstly, a single RGB-D camera is used to collect visual information in real time, and the YOLACT++ network is used to identify and segment the visual information to extract part of the important visual information. Secondly, we combine the three stages of depth recovery, depth optimization, and deep fusion to propose a three-dimensional position estimation method based on deep learning for joint coding of visual information. It can reduce the depth error caused by the depth measurement process, and the accurate 3D point values of the segmented image can be obtained directly. Finally, we propose a method based on the limited outlier adjustment of the cluster center distance to optimize the three-dimensional point values obtained above. It improves the real-time reconstruction accuracy and obtains the three-dimensional model of the object in real time. Experimental results show that this method only needs a single RGB-D camera, which is not only low cost and convenient to use, but also significantly improves the speed and accuracy of 3D reconstruction.

Neural networks and the anti-inflammatory effect of transcutaneous auricular vagus nerve stimulation in depression.

Transcutaneous auricular vagus nerve stimulation (taVNS) is a relatively non-invasive alternative treatment for patients suffering from major depressive disorder (MDD). It has been postulated that acupuncture may achieve its treatment effects on MDD through suppression of vagal nerve inflammatory responses. Our previous research established that taVNS significantly increases amygdala-dorsolateral prefrontal cortex connectivity, which is associated with a reduction in depression severity. However, the relationship between taVNS and the central/peripheral functional state of the immune system, as well as changes in brain neural circuits, have not as yet been elucidated. In the present paper, we outline the anatomic foundation of taVNS and emphasize that it significantly modulates the activity and connectivity of a wide range of neural networks, including the default mode network, executive network, and networks involved in emotional and reward circuits. In addition, we present the inflammatory mechanism of MDD and describe how taVNS inhibits central and peripheral inflammation, which is possibly related to the effectiveness of taVNS in reducing depression severity. Our review suggests a link between the suppression of inflammation and changes in brain regions/circuits post taVNS.

Convolutional Neural Networks-Based Object Detection Algorithm by Jointing Semantic Segmentation for Images.

In recent years, increasing image data comes from various sensors, and object detection plays a vital role in image understanding. For object detection in complex scenes, more detailed information in the image should be obtained to improve the accuracy of detection task. In this paper, we propose an object detection algorithm by jointing semantic segmentation (SSOD) for images. First, we construct a feature extraction network that integrates the hourglass structure network with the attention mechanism layer to extract and fuse multi-scale features to generate high-level features with rich semantic information. Second, the semantic segmentation task is used as an auxiliary task to allow the algorithm to perform multi-task learning. Finally, multi-scale features are used to predict the location and category of the object. The experimental results show that our algorithm substantially enhances object detection performance and consistently outperforms other three comparison algorithms, and the detection speed can reach real-time, which can be used for real-time detection.

Decompressive hemicraniectomy in patients with malignant middle cerebral artery infarction: A systematic review and meta-analysis.

BACKGROUND & PURPOSE: Malignant middle cerebral artery infarctions (mMCAI) are one of the most devastating ischemic strokes, with up to 80% mortality in non-surgically treated patients. With the publication of three European randomized controlled trials (RCTs), decompressive hemicraniectomy (DHC) was recommended in patients with mMCAI who are aged ≤ 60 years. Recently, three other RCTs enrolling patients aged > 60 years were published; thus, it is necessary to update the previous meta-analysis to re-evaluate the effects of DHC in mMCAI. METHODS: A systematic literature search of PubMed, EMBASE, and the Cochrane Library was conducted for published RCTs investigating the effects of DHC in mMCAI. Primary outcomes were mortality and major disability (modified Rankin Scale score: 4-5) among survivors. Secondary outcomes were death or major disability (mRS score > 3), and death or severe disability (mRS score > 4). Effect sizes were expressed in Peto odds ratio (Peto OR) with 95% confidence intervals. RESULTS: Six studies with 314 patients were subjected to meta-analysis. Data showed that DHC, significantly decreased mortality risk, death or major disability (mRS score > 3), and death or severe disability (mRS score > 4); but was associated with a slightly higher proportion of major disability (mRS score: 4-5) among survivors. There were no statistically significant age differences. CONCLUSIONS: Compared to conservative treatment, DHC significantly decreased mortality and improved functional outcome, with a non-significant increase in the proportion of survivors with major disability. Further studies are required for multidimensional evaluation of DHC for mMCAI.

Impact response of lightweight steel foam concrete composite slabs: Experimental, numerical and analytical studies.

This paper presents a study on the low-velocity impact response of lightweight steel foam concrete (LSFC) composite slabs. The LSFC composite slab consisted of a W-shaped steel plate, foam concrete and oriented strand board (OSB). Low-velocity impact tests on the LSFC composite slabs were conducted by employing an ultra-high heavy-duty drop hammer testing machine. The tests revealed the failure mode, impact force and displacement response of LSFC composite slabs. The effects of density and thickness of foam concrete and drop height on the peak impact force and energy absorption ratio were investigated. A finite element (FE) model was set up to predict the impact resistance of the LSFC composite slabs, and a good agreement between simulation and test results was achieved. In addition, an equivalent-single-degree-of-freedom (ESDOF) model was set up to predict the displacement response of the LSFC composite slabs under impact loading.

The safety and clinical outcomes of endovascular treatment versus microsurgical clipping of ruptured anterior communicating artery aneurysms: a 2-year follow-up, multicenter, observational study.

Background and objective: Current data on the optimal treatment modality for ruptured anterior communicating artery (AComA) aneurysms are limited. We conducted this multicenter retrospective study to evaluate the safety and clinical outcomes of endovascular treatment (EVT) and microsurgical clipping (MC) for the treatment of ruptured AComA patients. Methods: Patients with ruptured AComA aneurysms were screened from the Chinese Multicenter Cerebral Aneurysm Database. Propensity score matching (PSM) was used to adjust for baseline characteristic imbalances between the EVT and MC groups. The safety outcomes included total procedural complications, procedure-related morbidity/death and remedial procedure for complication. The primary clinical outcome was 2-year functional independence measured by the modified Rankin scale (mRS) score. Results: The analysis included 893 patients with ruptured AComA aneurysms (EVT: 549; MC: 346). PSM yielded 275 pairs of patients in the EVT and MC cohorts for comparison. Decompressive craniectomy being more prevalent in the MC group (19.3% vs. 1.5%, p < 0.001). Safety data revealed a lower rate of total procedural complications (odds ratio [OR] = 0.62, 95% CI 0.39-0.99; p = 0.044) in the EVT group and similar rates of procedure-related morbidity/death (OR = 0.91, 95% CI 0.48-1.73; p = 0.880) and remedial procedure for complication (OR = 1.35, 95% CI 0.51-3.69, p = 0.657) between the groups. Compared with that of MC patients, EVT patients had a greater likelihood of functional independence (mRS score 0-2) at discharge (OR = 1.68, 95% CI 1.14-2.50; p = 0.008) and at 2 years (OR = 1.89, 95% CI 1.20-3.00; p = 0.005), a lower incidence of 2-year all-cause mortality (OR = 0.54, 95% CI 0.31-0.93; p = 0.023) and a similar rate of retreatment (OR = 1.00, 95% CI 0.23-4.40; p = 1.000). Conclusion: Clinical outcomes after treatment for ruptured AComA aneurysms appear to be superior to those after treatment with MC, with fewer overall procedure-related complications and no increase in the retreatment rate. Additional studies in other countries are needed to verify these findings.

Treatment and outcome analysis of patients with ruptured distal anterior cerebral artery aneurysms: a multicenter real-world study.

Objective: To reveal the safety and efficacy of clipping and coiling in patients with ruptured distal anterior cerebral artery aneurysms (DACAA) and to calculate the risk factors affecting the two-year survival rate in follow-up patients. Methods: A retrospective study was conducted on the data of 140 patients (21 were lost to follow-up) with DACAA rupture who were treated by neurosurgery at 12 medical centers over a 2-year period, from January 2017 to December 2020. Univariate analysis was used to examine factors contributing to poor patient prognosis and to compare the prognosis of coiling and clipping treatments. Survival analysis was employed to compare survival rates between coiling and clipping, and risk factors affecting patient survival were analyzed using multivariate Cox regression analysis. Results: Out of 140 patients with ruptured DACAA, 80 (57.1%) were male, and 60 (42.9%) were female. A total of 111 (79.3%) patients were classified under Hunt-Hess scale grades I-III, while 95 (67.9%) were graded I-III according to the WFNs classification. Among them, 63 (45%) were treated with clipping, and 77 (55%) underwent coiling. Within 2 years of discharge from the hospital, 31 (59.6%) patients who underwent clipping and 54 (80.6%) who underwent coiling had a good prognosis. Multivariate Cox regression analysis revealed that only WFNs classification (I-III) was a protective factor influencing the 2-year survival of patients with ruptured DACAA. Conclusion: In the reality of medical practice, neurosurgeons are more likely to choose clipping as the treatment for cases with WFNs classification than or equal to III. There was no difference between clipping and coiling in the two-year prognosis at discharge. High priority should be given to DACAA cases with WFNs grading (I-III), as better outcomes can be achieved. The sample size will continue to be enlarged in the future to obtain more accurate findings. Abstracts for reviews, technical notes, and historical vignettes do not need to be separated into sections. They should begin with a clear statement of the paper's purpose followed by appropriate details that support the authors' conclusion(s).

Dopamine inhibits high-frequency stimulation-induced long-term potentiation of intrinsic excitability in CA1 hippocampal pyramidal neurons.

The efficiency of neural circuits is modified by changes not only in synaptic strength, but also in intrinsic excitability of neurons. In CA1 hippocampal pyramidal neurons, bidirectional changes in the intrinsic excitability are often presented after induction of synaptic long-term potentiation or depression. This plasticity of intrinsic excitability has been identified as a cellular correlate of learning. Besides, behavioral learning often involves action of reinforcement or rewarding mediated by dopamine (DA). Here, we examined how DA influences the intrinsic plasticity of CA1 hippocampal pyramidal neurons when high-frequency stimulation (HFS) was applied to Schaffer collaterals. The results showed that DA inhibits the decrease in rheobase and increase in mean firing rate of pyramidal neurons induced by HFS, and that this inhibition was abolished by the D1-like receptor antagonist SCH23390 but not by the D2-like receptor antagonist sulpiride. The results suggest that DA inhibits the potentiation of excitability induced by presynaptic HFS, and that this inhibition depends on the activation of D1-like receptors.

Spatial distribution characteristics and migration of microplastics in surface water, groundwater and sediment in karst areas: The case of Yulong River in Guilin, Southwest China.

Karst regions, due to their unique topography, may be more susceptible to contaminants such as microplastics from other ecosystems. However, few studies reported the occurrence of microplastics in karst areas. Here, we investigated the abundance of microplastics in surface water, sediments and groundwater. In addition, their spatial distribution characteristics and migration were also analyzed in a typical karst area, Yulong River, Guilin, China. Microplastic pollution was found in the sediments, surface water and especially groundwater in Yulong River. The abundance of microplastics was 0-4 items/L, 247-1708 items/kg and 0-4 items/L in surface water, sediments and groundwater, respectively. Microplastics in surface water and groundwater were fiber-based, while those in sediments were fragment-based. Polypropylene (PP), Polystyrene (PS) and Polyethylene terephthalate (PET) are dominant microplastic types in Yulong River. Moreover, the abundance of microplastic pollution in different functional areas ranked as follows: living area > agricultural area > landscape area. Clustering analysis showed that disposable tableware and plastic bags used in tourism activities might be the main source of microplastics. More importantly, our findings suggested that microplastics in groundwater could be the result of hydraulic exchange between groundwater and surface water in karst areas, rather than soil infiltration. These findings provided us with a further understanding of the pollution of microplastics in karst rivers.