Recall DNA methylation levels at low coverage sites using a CNN model in WGBS.

DNA methylation is an important regulator of gene transcription. WGBS is the gold-standard approach for base-pair resolution quantitative of DNA methylation. It requires high sequencing depth. Many CpG sites with insufficient coverage in the WGBS data, resulting in inaccurate DNA methylation levels of individual sites. Many state-of-arts computation methods were proposed to predict the missing value. However, many methods required either other omics datasets or other cross-sample data. And most of them only predicted the state of DNA methylation. In this study, we proposed the RcWGBS, which can impute the missing (or low coverage) values from the DNA methylation levels on the adjacent sides. Deep learning techniques were employed for the accurate prediction. The WGBS datasets of H1-hESC and GM12878 were down-sampled. The average difference between the DNA methylation level at 12× depth predicted by RcWGBS and that at >50× depth in the H1-hESC and GM2878 cells are less than 0.03 and 0.01, respectively. RcWGBS performed better than METHimpute even though the sequencing depth was as low as 12×. Our work would help to process methylation data of low sequencing depth. It is beneficial for researchers to save sequencing costs and improve data utilization through computational methods.

Deciphering Microbial Adaptation in the Rhizosphere: Insights into Niche Preference, Functional Profiles, and Cross-Kingdom Co-occurrences.

Rhizosphere microbial communities are to be as critical factors for plant growth and vitality, and their adaptive differentiation strategies have received increasing amounts of attention but are poorly understood. In this study, we obtained bacterial and fungal amplicon sequences from the rhizosphere and bulk soils of various ecosystems to investigate the potential mechanisms of microbial adaptation to the rhizosphere environment. Our focus encompasses three aspects: niche preference, functional profiles, and cross-kingdom co-occurrence patterns. Our findings revealed a correlation between niche similarity and nucleotide distance, suggesting that niche adaptation explains nucleotide variation among some closely related amplicon sequence variants (ASVs). Furthermore, biological macromolecule metabolism and communication among abundant bacteria increase in the rhizosphere conditions, suggesting that bacterial function is trait-mediated in terms of fitness in new habitats. Additionally, our analysis of cross-kingdom networks revealed that fungi act as intermediaries that facilitate connections between bacteria, indicating that microbes can modify their cooperative relationships to adapt. Overall, the evidence for rhizosphere microbial community adaptation, via differences in gene and functional and co-occurrence patterns, elucidates the adaptive benefits of genetic and functional flexibility of the rhizosphere microbiota through niche shifts.

Machine learning for phytopathology: from the molecular scale towards the network scale.

With the increasing volume of high-throughput sequencing data from a variety of omics techniques in the field of plant-pathogen interactions, sorting, retrieving, processing and visualizing biological information have become a great challenge. Within the explosion of data, machine learning offers powerful tools to process these complex omics data by various algorithms, such as Bayesian reasoning, support vector machine and random forest. Here, we introduce the basic frameworks of machine learning in dissecting plant-pathogen interactions and discuss the applications and advances of machine learning in plant-pathogen interactions from molecular to network biology, including the prediction of pathogen effectors, plant disease resistance protein monitoring and the discovery of protein-protein networks. The aim of this review is to provide a summary of advances in plant defense and pathogen infection and to indicate the important developments of machine learning in phytopathology.

Effector-GAN: prediction of fungal effector proteins based on pretrained deep representation learning methods and generative adversarial networks.

MOTIVATION: Phytopathogenic fungi secrete effector proteins to subvert host defenses and facilitate infection. Systematic analysis and prediction of candidate fungal effector proteins are crucial for experimental validation and biological control of plant disease. However, two problems are still considered intractable to be solved in fungal effector prediction: one is the high-level diversity in effector sequences that increases the difficulty of protein feature learning, and the other is the class imbalance between effector and non-effector samples in the training dataset. RESULTS: In our study, pretrained deep representation learning methods are presented to represent multiple characteristics of sequences for predicting fungal effectors and generative adversarial networks are adapted to create synthetic feature samples to address the data imbalance problem. Compared with the state-of-the-art fungal effector prediction methods, Effector-GAN shows an overall improvement in accuracy in the independent test set. AVAILABILITY AND IMPLEMENTATION: Effector-GAN offers a user-friendly interface to inspect potential fungal effector proteins (http://lab.malab.cn/~wys/webserver/Effector-GAN). The Python script can be downloaded from http://lab.malab.cn/~wys/gitlab/effector-gan. SUPPLEMENTARY INFORMATION: Supplementary data are available at Bioinformatics online.

LINC00963-FOSB-mediated transcription activation of UBE3C enhances radioresistance of breast cancer cells by inducing ubiquitination-dependent protein degradation of TP73.

BACKGROUND: The ubiquitin protein ligase E3C (UBE3C) has been reported to play an oncogenic role in breast cancer (BRCA). This work further investigates the effect of UBE3C on the radioresistance of BRCA cells. METHODS: Molecules linking to radioresistance in BRCA were identified by analyzing two GEO datasets, GSE31863 and GSE101920. UBE3C overexpression or knockdown was induced in parental or radioresistant BRCA cells, followed by irradiation treatment. The malignant properties of cells in vitro, and the growth and metastatic activity of cells in nude mice, were analyzed. Downstream target proteins, as well as upstream transcriptional regulators of UBE3C, were predicted by bioinformatics tools. Molecular interactions were confirmed by immunoprecipitation and immunofluorescence assays. Furthermore, artificial alterations of TP73 and FOSB were induced in the BRCA cells for functional rescue assays. RESULTS: According to bioinformatics analyses, UBE3C expression was linked to radioresistance in BRCA. UBE3C knockdown in radioresistant BRCA cells reduced while its overexpression in parental BRCA cells increased the radioresistance of cells in vitro and in vivo. UBE3C, which induced ubiquitination-dependent protein degradation of TP73, was transcriptionally activated by FOSB. The radioresistance of cancer cells was blocked by TP73 overexpression or FOSB knockdown. Additionally, LINC00963 was found to be responsible for the recruitment of FOSB to the UBE3C promoter for transcription activation. CONCLUSION: This work demonstrates that LINC00963 induces nuclear translocation of FOSB and the consequent transcription activation of UBE3C, which enhances radioresistance of BRCA cells by inducing ubiquitination-dependent protein degradation of TP73.

Clusterin is closely associated with adipose tissue insulin resistance.

AIMS: Clusterin (encoded by CLU) is a novel adipokine. Serum clusterin levels were elevated in populations with obesity and diabetes. Adipose tissue insulin resistance (Adipo-IR) is proposed as an early metabolic defect that precedes systemic insulin resistance. Herein, we aimed to investigate the relationship between serum clusterin levels and Adipo-IR. CLU expression in human abdominal adipose tissues and clusterin secretion in human adipocytes was also explored. MATERIALS AND METHODS: A total of 201 participants (aged 18-62 years, 139 of whom were obese) were recruited. Enzyme-linked immunosorbent assay was used to measure serum clusterin levels. Adipo-IR was calculated from the product of fasting free fatty acids and fasting insulin levels. Transcriptome sequencing of abdominal visceral adipose tissue (VAT) and subcutaneous adipose tissue (SAT) was performed. Human adipocytes were used to detect the secretion of clusterin. RESULTS: Serum clusterin levels were independently associated with Adipo-IR after adjusting for several confounding factors (standardised ß = 0.165, p = 0.021). CLU expression in VAT and SAT was associated with obesity-related metabolic risk factors. Higher CLU expression in VAT was accompanied by an increase in collagen accumulation. Clusterin secretion in differentiated human adipocytes was stimulated by insulin and inhibited by rosiglitazone. CONCLUSIONS: Clusterin is strongly associated with Adipo-IR. Serum clusterin may function as an effective indicator of adipose tissue insulin resistance.

Neutrophil-to-lymphocyte and platelet-to-lymphocyte ratios as prognostic biomarkers in limited-stage small-cell lung cancer: a meta-analysis.

Background: The role of neutrophil-to-lymphocyte ratio (NLR) and platelet-to-lymphocyte ratio (PLR) as prognostic markers in limited-stage small-cell lung cancer (LS-SCLC) remains controversial. Methods: Using pooled hazard ratios (HR) with 95% CIs, we assessed the correlation of pre-treatment NLR and PLR with overall survival (OS) and progression-free survival (PFS) in LS-SCLC. Publication bias was assessed by Begg's and Egger's tests. Results: Ten studies were enrolled in our meta-analysis. Pooled analyses showed that pre-treatment high NLR was significantly associated with poor OS (HR: 1.80) and PFS (HR: 1.69) in LS-SCLC patients. High pre-treatment PLR was also associated with shorter OS (HR: 1.52) and PFS (HR: 1.39) in LS-SCLC patients. Conclusion: Our meta-analysis suggests that high pre-treatment NLR or PLR may be negatively related to OS and PFS in LS-SCLC.

Metagenomics reveals the self-recovery and risk of antibiotic resistomes during carcass decomposition of wild mammals.

Animal carcass decomposition may bring serious harm to the environment, including pathogenic viruses, toxic gases and metabolites, and antibiotic resistance genes (ARGs). However, how wild mammal corpses decomposition influence and change ARGs in the environment has less explored. Through metagenomics, 16S rRNA gene sequencing, and physicochemical analysis, this study explored the succession patterns, influencing factors, and assembly process of ARGs and mobile genetic elements (MGEs) in gravesoil during long-term corpse decomposition of wild mammals. Our results indicate that the ARG and MGE communities related to wildlife corpses exhibited a pattern of differentiation first and then convergence. Different from the farmed animals, the decomposition of wild animals first reduced the diversity of ARGs and MGEs, and then recovered to a level similar to that of the control group (untreated soil). ARGs and MGEs of the gravesoil are mainly affected by deterministic processes in different stages. MGEs and bacterial community are the two most important factors affecting ARGs in gravesoil. It is worth noting that the decomposition of wild animal carcasses enriched different high-risk ARGs at different stages (bacA, mecA and floR), which have co-occurrence patterns with opportunistic pathogens (Comamonas and Acinetobacter), thereby posing a great threat to public health. These results are of great significance for wildlife corpse management and environmental and ecological safety.

prPred-DRLF: Plant R protein predictor using deep representation learning features.

Plant resistance (R) proteins play a significant role in the detection of pathogen invasion. Accurately predicting plant R proteins is a key task in phytopathology. Most plant R protein predictors are dependent on traditional feature extraction methods. Recently, deep representation learning methods have been successfully applied in solving protein classification problems. Motivated by this, we propose a new computational approach, called prPred-DRLF, which uses deep representation learning feature models to encode the amino acids as numerical vectors. The results show that the fused features of bidirectional long short-term memory (BiLSTM) embedding and unified representation (UniRep) embedding have a better performance than other features for plant R protein identification using a light gradient boosting machine (LGBM) classifier. The model was evaluated using an independent test achieving an accuracy of 0.956, F1-score of 0.933, and area under the receiver operating characteristic (ROC) curve (AUC) of 0.997. Meanwhile, compared with the state-of-the-art prPred and HMMER method, prPred-DRLF shows an overall improvement in accuracy, F1-score, AUC, and recall. prPred-DRLF is a higher-performance plant R protein prediction tool based on two kinds of deep representation learning technologies and offers a user-friendly interface for inspecting possible plant R proteins. We hope that prPred-DRLF will become a useful tool for biological research. A user-friendly webserver for prPred-DRLF is freely accessible at http://lab.malab.cn/soft/prPred-DRLF. The Python script can be downloaded from https://github.com/Wangys-prog/prPred-DRLF.

Comparative genome analysis of plant ascomycete fungal pathogens with different lifestyles reveals distinctive virulence strategies.

BACKGROUND: Pathogens have evolved diverse lifestyles and adopted pivotal new roles in both natural ecosystems and human environments. However, the molecular mechanisms underlying their adaptation to new lifestyles are obscure. Comparative genomics was adopted to determine distinct strategies of plant ascomycete fungal pathogens with different lifestyles and to elucidate their distinctive virulence strategies. RESULTS: We found that plant ascomycete biotrophs exhibited lower gene gain and loss events and loss of CAZyme-encoding genes involved in plant cell wall degradation and biosynthesis gene clusters for the production of secondary metabolites in the genome. Comparison with the candidate effectome detected distinctive variations between plant biotrophic pathogens and other groups (including human, necrotrophic and hemibiotrophic pathogens). The results revealed the biotroph-specific and lifestyle-conserved candidate effector families. These data have been configured in web-based genome browser applications for public display ( http://lab.malab.cn/soft/PFPG ). This resource allows researchers to profile the genome, proteome, secretome and effectome of plant fungal pathogens. CONCLUSIONS: Our findings demonstrated different genome evolution strategies of plant fungal pathogens with different lifestyles and explored their lifestyle-conserved and specific candidate effectors. It will provide a new basis for discovering the novel effectors and their pathogenic mechanisms.

The mediating role of the visceral fat area in the correlation between the serum osteocalcin levels and a prolonged QTc interval.

AIMS: Osteocalcin, a bone-derived factor, could be a feasible marker for metabolic disorders and adverse cardiovascular outcomes. This study aimed to explore the correlation between serum osteocalcin levels and correct QT interval (QTc) interval prolongation, a risk factor of cardiac morbidity and mortality. METHODS: We recruited 1210 subjects (age range: 26-80 years) in communities in Shanghai. Serum osteocalcin levels were determined using an electrochemiluminescence immunoassay. The QTc interval was measured using a 12-lead electrocardiogram and was calculated by the Bazett formula. A prolonged QTc interval was defined as QTc > 440 ms. Visceral fat area (VFA) was assessed by magnetic resonance imaging. A VFA of 80 cm2 was applied as a cut-off point for central obesity. RESULTS: Subjects with diabetes, overweight/obesity, or central obesity had significantly lower serum osteocalcin levels than those without (all P < 0.01). In subjects with a normal QTc interval, QTc interval lengthening accompanied decreasing osteocalcin levels (Pfor trend = 0.033), and the decline was more obvious in subjects with a prolonged QTc interval (Pfor trend = 0.022). Serum osteocalcin levels were correlated with the QTc interval (standardized ß = -0.082, P = 0.005). Neither diabetes nor overweight/obesity was correlated with the QTc interval, whereas central obesity was positively correlated (P = 0.032). In addition, the correlation between osteocalcin levels and the QTc interval was attenuated when central obesity was included in the model simultaneously (standardized ß = -0.075, P = 0.010). Mediation analysis revealed that VFA played a mediating role in the aforementioned correlation, and the estimated percentage of the total effect mediated by VFA was 20.9% (P = 0.007). CONCLUSIONS: VFA partially mediated the inverse correlation between the serum osteocalcin levels and QTc interval, suggesting that improving fat metabolism may be a mechanism by which osteocalcin protects against cardiovascular diseases.

Ceratocystis fimbriata Employs a Unique Infection Strategy Targeting Peltate Glandular Trichomes of Sweetpotato (Ipomoea batatas) Plants.

The infection processes of Ceratocystis fimbriata BMPZ13 (BMPZ13) was elucidated on vegetative tissues of sweetpotato plants employing light and scanning electron microscopy. Vegetative tissues infected with C. fimbriata BMPZ13 by either wounding or nonwounding inoculation methods developed typical disease symptoms, establishing black rot in stems and necrosis on buds, young leaves, and stems of sprouts, in addition to wilt on leaves and shoot cuttings, typical of vascular associated diseases. The runner hyphae of C. fimbriata BMPZ13 formed from germinated conidia were able to directly penetrate the epidermal cuticle for initial infection and invade sweetpotato peltate glandular trichomes, specialized secretory structures to store and secrete metabolites. A two-step biotrophic phase was observed with nonwounding inoculation on leaves and stems, featuring both intercellular and intracellular invasive hyphae, with the latter found within living cells of the leaf epidermis. Subsequent to the biotrophic phase was a necrotrophic phase displaying cell death in infected leaves and veins. Additionally, this cell death was an iron-associated ferroptosis, supporting the notion that iron is involved in the necrotrophic phase of C. fimbriata BMPZ13 infection. Significantly, we establish that C. fimbriata employs a unique infection strategy: the targeting of peltate glandular trichomes. Collectively, our findings show that C. fimbriata is a plant fungal pathogen with a hemibiotrophic infection style in sweetpotato vegetative tissues.

Binding and detoxification of chlorpyrifos by lactic acid bacteria on rice straw silage fermentation.

This investigation examined the reduction of pesticide residues on straw inoculated with lactic acid bacteria (LAB) during ensiling. Lactobacillus casei WYS3 was isolated from rice straw that contained pesticide residues. Non-sterilized rice straw, which was inoculated with L. casei WYS3, showed increased removal of chlorpyrifos after ensiling, compared with rice straw that was not inoculated with L. casei WYS3 or sterilized rice straw. In pure culture, these strains can bind chlorpyrifos as indicated by high-performance liquid chromatography analysis. Viable L. casei WYS3 was shown to bind 33.3-42% of exogenously added chlorpyrifos. These results are similar to those of acid-treated cells but less than those of heat-treated cells, which were found to bind 32.0% and 77.2% of the added chlorpyrifos respectively. Furthermore, gas chromatography-mass spectrometry analysis determined that L. casei WYS3 detoxified chlorpyrifos via P-O-C cleavage. Real-time polymerized chain reaction analysis determined that organophosphorus hydrolase gene expression tripled after the addition of chlorpyrifos to LAB cultures, compared with the control group (without chlorpyrifos). This paper highlights the potential use of LAB starter cultures for the detoxification and removal of chlorpyrifos residues in the environment.

Application and Comparison of Machine Learning and Database-Based Methods in Taxonomic Classification of High-Throughput Sequencing Data.

The advent of high-throughput sequencing technologies has not only revolutionized the field of bioinformatics but has also heightened the demand for efficient taxonomic classification. Despite technological advancements, efficiently processing and analyzing the deluge of sequencing data for precise taxonomic classification remains a formidable challenge. Existing classification approaches primarily fall into two categories, database-based methods and machine learning methods, each presenting its own set of challenges and advantages. On this basis, the aim of our study was to conduct a comparative analysis between these two methods while also investigating the merits of integrating multiple database-based methods. Through an in-depth comparative study, we evaluated the performance of both methodological categories in taxonomic classification by utilizing simulated data sets. Our analysis revealed that database-based methods excel in classification accuracy when backed by a rich and comprehensive reference database. Conversely, while machine learning methods show superior performance in scenarios where reference sequences are sparse or lacking, they generally show inferior performance compared with database methods under most conditions. Moreover, our study confirms that integrating multiple database-based methods does, in fact, enhance classification accuracy. These findings shed new light on the taxonomic classification of high-throughput sequencing data and bear substantial implications for the future development of computational biology. For those interested in further exploring our methods, the source code of this study is publicly available on https://github.com/LoadStar822/Genome-Classifier-Performance-Evaluator. Additionally, a dedicated webpage showcasing our collected database, data sets, and various classification software can be found at http://lab.malab.cn/~tqz/project/taxonomic/.

Associations of Serum Uric Acid to High-Density Lipoprotein Cholesterol Ratio with Trunk Fat Mass and Visceral Fat Accumulation.

Background: It has been reported recently that the ratio of uric acid to high-density lipoprotein cholesterol (UHR) is correlated with several metabolic disorders. The present study aimed to investigate the associations of UHR with body fat content and distribution. Methods: This study enrolled 300 participants (58 men and 242 women) aged 18 to 65 years. The levels of serum uric acid and high-density lipoprotein cholesterol were measured by standard enzymatic methods. The overall fat content and segmental fat distribution were assessed with an automatic bioelectrical impedance analyzer. In the population with obesity, the visceral fat area (VFA) and subcutaneous fat area (SFA) were measured using magnetic resonance imaging. Results: Among the study population, 219 individuals (73.0%) were with obesity. The median level of UHR in individuals with obesity was 33.7% (26.2% - 45.9%), which was significantly higher than that in those without obesity [22.6% (17.0% - 34.4%), P < 0.01]. UHR was positively associated with overall fat content and segmental fat distribution parameters (all P < 0.01). In multivariate linear regression analysis, compared with body mass index, waist circumference was more closely associated with UHR (standardized ß = 0.427, P < 0.001) after adjusting for confounding factors. Additionally, total fat mass (standardized ß = 0.225, P = 0.002) and trunk fat mass (standardized ß = 0.296, P = 0.036) were more closely linked to UHR than total fat-free mass and leg fat mass, respectively. In the population with obesity, VFA was independently correlated with UHR (P < 0.01), while SFA was not associated with UHR. Conclusion: UHR was significantly associated with overall fat content and trunk fat accumulation. In the population with obesity, UHR was positively associated with VFA. Attention should be paid to the role of excessive trunk fat mass in the relationship between UHR and metabolic disorders.

Toxicity of chromium to wheat (Triticum aestivum L.) in two soils: influence of soil properties and chromium form.

The toxicity of Cr to plants depends on Cr form and soil properties. Currently, the phytotoxicity differences of Cr(VI) and Cr(III) in different soils are not clear. In this study, the toxicity of Cr(VI) and Cr(III) to root growth and root morphology of wheat (Triticum aestivum L.) were compared in Shandong fluvo-aquic soil (SD soil) and Jiangxi red soil (JX soil) that is differing in soil properties. The toxicity thresholds of Cr(VI) and Cr(III) on wheat root elongation were determined by fitting the dose-effect curves. Results showed that the 10% and 50% root length inhibitory concentrations (EC10 and EC50) of Cr(III) were 53.1 and 125 times of Cr(VI) in SD soil and 8.11 and 1.36 times of Cr(VI) in JX soil, indicating that Cr(VI) was more toxic to wheat roots than Cr(III) in both soils and the toxicity discrepancy of the two forms of Cr was more prominent in SD soil. Cr(VI) exhibited higher toxicity in SD soil (alkaline) than in JX soil (acidic), whereas Cr(III) showed the opposite pattern. In addition, the ethylene diamine tetraacetic acid extractable Cr (EDTA-Cr) concentrations in soils were correlated well with the relative wheat root elongation (R2=0.854, P<0.01), indicating that soil EDTA-Cr concentration can be used as a predictor of Cr phytotoxicity. Both Cr(VI) and Cr(III) showed significant biphasic dose effects on wheat root morphology (root length, root surface area, root volume, and root tip number) in JX soil. These findings are helpful for the risk evaluation of Cr contamination in agricultural soils.

Genome-wide analysis of the LBD family in rice: Gene functions, structure and evolution.

Recent evidence suggests that LATERAL ORGAN BOUNDARIES DOMAIN (LBD) proteins are involved in different developmental processes of plants. Although the roles of LBD proteins in root development, leaf development and plant defense have been well summarized, their functional diversity and regulation mechanisms are still unclear. One of the reasons for the above problems is the lack of selection and classification of functional protein features of LBD genes. Combined with the existing research results, we found that LBD genes have similar features and mechanics and tend to be in the same phylogenetic branch. Research on the function of the LBD gene can expand our understanding of the diversity and function of LBD proteins. Therefore, to fully understand this large family, it is necessary to review functional studies through in-depth phylogenetic analysis of more genome-available species.

A retrospective analysis of prognostic factors and treatment choices in small cell lung cancer with liver metastasis.

Background: Small cell lung cancer (SCLC) is characterized by high aggressiveness and early dissemination, with the liver being the most common site of metastasis. Although it has been established that the prognosis for SCLC with liver metastasis is exceedingly poor, comprehensive data on clinical features, prognostic factors, treatment options, and outcomes of this patient population remain limited. This retrospective study aims to examine the clinicopathological features and current treatment landscape and to identify prognostic factors associated with SCLC with liver metastasis in real-world settings. Methods: We conducted a retrospective analysis of data on SCLC patients with liver metastasis at initial diagnosis between January 1, 2013, and January 1, 2022. Kaplan-Meier analysis and log-rank tests were employed to estimate the overall survival (OS) and progression-free survival (PFS). Cox regression models were utilized to identify independent prognostic factors. Results: A total of 349 patients were included in the study, with 97.7% of patients exhibiting pure SCLC and 42.4% of patients presenting with concomitant bone metastasis. Approximately one-fourth of the patients had metastases in ≥3 organs, and 18.9% of patients had an Eastern Cooperative Oncology Group performance status (ECOG PS) ≥2. The median OS was 10.97 months (95% CI: 9.88-12.06) for those who received first-line therapy (n=286). Of these, 263 patients were treated with chemotherapy, showing a median OS of 11.37 months. Furthermore, 43.8% of patients received second-line treatment, and 81 patients proceeded to third-line treatment. ECOG PS ≥2 and mixed-SCLC were identified as independent adverse prognostic factors in SCLC with liver metastasis, whereas treatments including systemic treatment alone or in combination with local radiotherapy were associated with better prognoses. Conclusions: This retrospective study substantiated that ECOG PS ≥2 and mixed SCLC are independent predictors of poor prognosis for SCLC with liver metastasis. Additionally, different treatment strategies can improve the survival of this patient population, with chemotherapy currently being the main treatment option.

Associations between serum clusterin levels and non-alcoholic fatty liver disease.

Objective: Clusterin is closely correlated with insulin resistance and its associated comorbidities. This study aimed to investigate the correlation between serum clusterin levels and non-alcoholic fatty liver disease (NAFLD) and further explore the mediating role of insulin resistance in this relationship. Methods: This study enrolled 195 inpatients (41 males and 154 females) aged 18-61 years. Twenty-four patients were followed up for 12 months after bariatric surgery. Serum clusterin levels were measured using a sandwich enzyme-linked immunosorbent assay. Fatty liver disease was diagnosed on the basis of liver ultrasonography. The fatty liver index (FLI) was calculated to quantify the degree of hepatic steatosis. The mediating role of homeostasis model assessment-insulin resistance (HOMA-IR) was assessed using mediation analysis. Results: Participants with NAFLD had significantly higher serum clusterin levels than those without NAFLD (444.61 (325.76-611.52) mg/L vs 294.10 (255.20-373.55) mg/L, P < 0.01). With increasing tertiles of serum clusterin levels, the prevalence of NAFLD displayed an upward trend (P < 0.01). Multivariate linear regression analysis showed that serum clusterin levels were a positive determinant of FLI (standardized ß = 0.271, P < 0.001) after adjusting for multiple metabolic risk factors. Serum clusterin levels significantly decreased after bariatric surgery (298.77 (262.56-358.10) mg/L vs 520.55 (354.94-750.21) mg/L, P < 0.01). In the mediation analysis, HOMA-IR played a mediating role in the correlation between serum clusterin levels and FLI; the estimated percentage of the total effect was 17.3%. Conclusion: Serum clusterin levels were associated with NAFLD. In addition, insulin resistance partially mediated the relationship between serum clusterin levels and FLI.