OBJECTIVE To study the ameliorative effect and mechanism of Forsythia suspensa-Lonicera japonica herb pair on acute lung injury （ALI） based on serum exosomal microRNA （miRNA）. METHODS The rats were randomly divided into a blank group （normal saline）， model group （nomal saline）， and F. suspensa-L. japonica herb pair group （2.55 g/kg）， with 10 rats in each group. Except for the blank group， the other groups were used to establish an ALI model by intratracheal dripping of 5 mg/ mL lipopolysaccharides. After modeling， each group was given relevant medicine/normal saline intragastrically， once a day， for 3 consecutive days. After the last medication， the pathological status of lung tissue was observed； lung wet-to-dry weight ratio and leukocyte counts in bronchoalveolar lavage fluid （BALF） were determined. The levels of inflammatory factors [tumor necrosis factor-α（TNF-α）， interleukin-1β （IL-1β）， IL-10] in BALF were determined. Exosomes were isolated from rat serum， and high- throughput sequencing technology was employed to screen differentially expressed miRNA within the exosomes， followed by Kyoto Encyclopedia of Genes and Genomes （KEGG） enrichment analysis. Based on the screened differentially expressed miRNA and the enriched KEGG pathways， in vitro cellular experiments were conducted for validation. RESULTS The animal experimental results demonstrated that after intervention with the F. suspensa-L. japonica herb pair， the wet-to-dry weight ratio， the number of leukocytes in BALF， as well as the levels of TNF-α and IL-1β in BALF of ALI rats were all significantly reduced （P＜0.01）， while the level of IL-10 was significantly increased （P＜0.01）. The results of high-throughput sequencing experiments revealed that the F. suspensa-L. japonica herb pair could significantly up-regulate the expressions of miR-345-3p， miR-194-5p， miR-653-5p， and others in exosomes. Among them， the KEGG pathways involved in the target genes of differentially expressed miRNA included the hypoxia-inducible factor-1（HIF-1） signaling pathway， among others. The results of cellular E-mail：huang.haiying@126.com validation experiments showed that overexpressed miR-345-3p could significantly elevate the level of IL-10 in the cell supernatant （P＜0.01）， while significantly reducing the levels of TNF-α and IL-1β in the cell supernatant， as well as the mRNA and protein expression levels of protein kinase B1， phosphatidylinositol 3- kinase， and HIF-1α （P＜0.01）. CONCLUSIONS F. suspensa-L. japonica herb pair can alleviate inflammatory responses and thereby exert a therapeutic effect in improving ALI by up-regulating the expression of miR-345-3p in serum exosomes and inhibiting the activity of the HIF-1 signaling pathway.

Diabetic nephropathy （DKD）， as a common microvascular complication of diabetes mellitus （DM）， is a major cause of end-stage renal disease （ESRD）. Its clinical manifestations include increased urinary protein excretion， thickening of the glomerular basement membrane， and renal tubulointerstitial fibrosis. The pathogenesis of DKD is complex and involves multiple factors， including disordered glucose metabolism， hemodynamic alterations， and oxidative stress. Although modern medical approaches can alleviate certain symptoms， they still have limitations such as insufficient therapeutic targeting and prominent adverse effects. The transforming growth factor-β/Smad （TGF-β/Smad） signaling pathway is not only a tissue fibrosis pathway that has attracted considerable attention in recent years， but also regulates multiple protein molecules， including the glomerular podocyte slit diaphragm protein Podocin， interleukin-1β （IL-1β）， and superoxide dismutase （SOD）， thereby participating in various pathological processes and ultimately mediating renal injury. Flavonoid compounds， owing to their sustained pharmacological effects， broad spectrum of action， and high safety profile， have become ideal candidates for targeted therapy research in DKD. Existing studies have shown that these compounds can exert inhibitory effects on renal fibrosis， alleviate inflammatory responses， protect podocytes， and reduce oxidative stress by regulating the interactions between the TGF-β/Smad signaling pathway and the aforementioned protein molecules， thereby maintaining renal structure and function， reducing proteinuria， and significantly improving DKD lesions. This review briefly outlines the composition and functions of the TGF-β/Smad signaling pathway， elucidates the mechanisms by which this pathway regulates DKD， and focuses on summarizing major studies from the past decade on flavonoid-based interventions in DKD through targeted inhibition of the TGF-β/Smad signaling pathway. Furthermore， it discusses the considerable therapeutic potential of flavonoids in the treatment of this disease， aiming to provide a scientific basis for future clinical prevention and treatment of DKD and to promote the development of targeted drugs.

Parkinson’s disease (PD) is a progressive neurodegenerative disorder characterized by the selective loss of dopaminergic neurons in the substantia nigra pars compacta and the pathological accumulation ofα‑synuclein. Although extensive progress has been made in elucidating its pathogenesis, current therapeutic approaches remain largely symptomatic, and effective disease-modifying treatments are still unavailable. Increasing evidence indicates that PD is driven by the interaction of multiple pathological processes, including neuroinflammation, iron homeostasis dysregulation and ferroptosis, endoplasmic reticulum (ER) stress, mitochondrial dysfunction, oxidative stress, and impaired protein homeostasis, which together contribute to neuronal vulnerability and degeneration. Fibroblast growth factors (FGFs) comprise a family of 22 ligands that play important roles in neural development, stress responses, metabolic regulation, and the maintenance of nervous system homeostasis. Recent studies have shown that several FGF family members, such as FGF1, FGF2, FGF9, and FGF21, exert neuroprotective effects in cellular and animal models of PD. These effects include the regulation of inflammatory responses, oxidative stress, iron homeostasis, cellular stress adaptation, and neuronal survival. Compared with therapeutic strategies targeting a single pathogenic pathway, FGFs appear to influence multiple disease-related processes, suggesting their potential relevance to the complex pathophysiology of PD. Experimental evidence indicates that altered FGF signaling may contribute to dopaminergic neuron dysfunction through the coordinated regulation of several interconnected mechanisms. FGFs have been reported to modulate neuroinflammation by affecting the activation of microglia and astrocytes, thereby influencing the inflammatory environment in the central nervous system. In addition, FGFs are involved in the regulation of iron homeostasis and ferroptosis, partly through antioxidant signaling pathways associated with NRF2, SLC7A11, and GPX4. Moreover, FGFs can alleviate ER stress and mitochondrial dysfunction by activating intracellular signaling pathways such as PI3K/AKT, AMPK-PGC-1α, as well as SIRT1-dependent programs, which support cellular energy metabolism and redox balance. Recent advances in single-cell and spatial transcriptomic studies further suggest that FGF signaling is not limited to neuron-intrinsic mechanisms but also involves interactions among different glial cell types. Altered FGF ligand-receptor communication between astrocytes and oligodendrocytes has been observed in PD models and is associated with increased susceptibility of dopaminergic neurons to oxidative stress and ferroptosis. These findings indicate that the biological effects of FGFs are influenced by cell type and disease stage and may vary under different pathological conditions. In this review, we summarize recent progress in understanding the roles of FGF family members in PD, with a focus on their involvement in iron homeostasis dysregulation and ferroptosis, neuroinflammation, cellular stress responses, and neuronal protection and regeneration. By integrating current evidence, this review aims to provide a clearer understanding of how FGFs participate in PD pathogenesis and to offer a theoretical basis for future studies exploring their potential value in disease-modifying therapeutic strategies.

ObjectiveThe essence of syndrome manifestation recognition in traditional Chinese medicine (TCM) is to infer the body’s latent pathogenesis state from clinical observational information, rather than to perform simple label matching. However, previous studies have largely modeled this task as syndrome pattern classification within a fixed label space, which does not adequately reflect the cognition process of TCM syndrome differentiation centered on pathogenesis reasoning, and is also insufficient to capture the openness, semantic variability, and cross-disease reusability of syndrome manifestation expression. This study aimed to investigate whether introducing pathogenesis reasoning chain-of-thought (PR-CoT) supervision into large language models (LLMs) could improve the quality and cognitive consistency of syndrome manifestation recognition and support cross-disease transfer. MethodsSyndrome manifestation recognition was formulated as a conditional generation task under the framework of clinical observational information (X)→pathogenesis structure (Z)→syndrome pattern output (Y), where Z serves as an explicit intermediate structural variable linking the clinical evidence and syndrome judgment. Within this framework, a PR-CoT-supervised dataset for syndrome manifestation recognition was constructed based on medical case records of spleen-stomach disorders. After preprocessing, information extraction, manual proofreading, and data cleaning, the dataset comprised 4 800 training cases, 400 development cases, and 400 test cases. Each sample was annotated with a structured PR-CoT consisting of three progressive levels: clinical information summarization, comprehensive pathogenesis analysis, and syndrome pattern output. Supervised fine-tuning was conducted on open-source LLMs, with an end-to-end model serving as the baseline. Qwen3-32B was used as the primary experimental model, and Qwen3-14B as the scale comparison model. A progressive multidimensional evaluation framework was further established, comprising a structural parsing level, a semantic similarity level, and an expert blind review level. At the structural parsing level, syndrome pattern expressions were decomposed into structural elements and evaluated using Precision, Recall, F1 score, and Jaccard similarity. At the semantic similarity level, independent LLMs scored the theoretical proximity between predicted and reference syndrome patterns. At the expert blind review level, three TCM experts independently evaluated model outputs on two dimensions: syndrome differentiation consistency and terminology standardization of syndrome patterns. In addition, zero-shot cross-disease transfer evaluation was conducted on gynecological and heart-system disorder test sets. ResultsAt the structural parsing level, PR-CoT supervision did not lead to a stable improvement in the element-wise overlap of syndrome pattern structural components. Compared with the corresponding baselines, neither Qwen3-32B nor Qwen3-14B showed consistent advantages in structural matching metrics after the introduction of PR-CoT supervision. In contrast, at the semantic similarity level, PR-CoT supervision produced stable positive gains across different model scales and evaluation systems. The average semantic score of Qwen3-32B increased from 6.425 8 in the baseline model to 6.585 0 after PR-CoT supervision, and that of Qwen3-14B increased from 5.870 0 to 5.964 2. At the expert blind review level, the overall score of Qwen3-32B (PR-CoT) was 7.026 0±0.107 7, higher than 6.416 3±0.288 9 for its baseline. In zero-shot cross-disease testing, the PR-CoT model still showed advantages in semantic evaluation and expert evaluation on both gynecological and heart-system disorder test sets, indicating a certain degree of transferability. ConclusionThe benefits of PR-CoT supervision are mainly reflected in TCM semantic consistency and clinical plausibility, rather than in improved hard matching of structural elements. These findings support understanding syndrome manifestation recognition as a process of generating and expressing latent pathogenesis structures, rather than as a classification task within a traditional fixed label space. By introducing pathogenesis reasoning as an explicit intermediate structure into the modeling process and combining it with a progressive multidimensional evaluation framework, this study provides a methodological pathway for intelligent TCM syndrome differentiation that integrates theoretical alignment, interpretability, and multi-level evaluation.

ObjectiveThe essence of syndrome manifestation recognition in traditional Chinese medicine (TCM) is to infer the body’s latent pathogenesis state from clinical observational information, rather than to perform simple label matching. However, previous studies have largely modeled this task as syndrome pattern classification within a fixed label space, which does not adequately reflect the cognition process of TCM syndrome differentiation centered on pathogenesis reasoning, and is also insufficient to capture the openness, semantic variability, and cross-disease reusability of syndrome manifestation expression. This study aimed to investigate whether introducing pathogenesis reasoning chain-of-thought (PR-CoT) supervision into large language models (LLMs) could improve the quality and cognitive consistency of syndrome manifestation recognition and support cross-disease transfer. MethodsSyndrome manifestation recognition was formulated as a conditional generation task under the framework of clinical observational information (X)→pathogenesis structure (Z)→syndrome pattern output (Y), where Z serves as an explicit intermediate structural variable linking the clinical evidence and syndrome judgment. Within this framework, a PR-CoT-supervised dataset for syndrome manifestation recognition was constructed based on medical case records of spleen-stomach disorders. After preprocessing, information extraction, manual proofreading, and data cleaning, the dataset comprised 4 800 training cases, 400 development cases, and 400 test cases. Each sample was annotated with a structured PR-CoT consisting of three progressive levels: clinical information summarization, comprehensive pathogenesis analysis, and syndrome pattern output. Supervised fine-tuning was conducted on open-source LLMs, with an end-to-end model serving as the baseline. Qwen3-32B was used as the primary experimental model, and Qwen3-14B as the scale comparison model. A progressive multidimensional evaluation framework was further established, comprising a structural parsing level, a semantic similarity level, and an expert blind review level. At the structural parsing level, syndrome pattern expressions were decomposed into structural elements and evaluated using Precision, Recall, F1 score, and Jaccard similarity. At the semantic similarity level, independent LLMs scored the theoretical proximity between predicted and reference syndrome patterns. At the expert blind review level, three TCM experts independently evaluated model outputs on two dimensions: syndrome differentiation consistency and terminology standardization of syndrome patterns. In addition, zero-shot cross-disease transfer evaluation was conducted on gynecological and heart-system disorder test sets. ResultsAt the structural parsing level, PR-CoT supervision did not lead to a stable improvement in the element-wise overlap of syndrome pattern structural components. Compared with the corresponding baselines, neither Qwen3-32B nor Qwen3-14B showed consistent advantages in structural matching metrics after the introduction of PR-CoT supervision. In contrast, at the semantic similarity level, PR-CoT supervision produced stable positive gains across different model scales and evaluation systems. The average semantic score of Qwen3-32B increased from 6.425 8 in the baseline model to 6.585 0 after PR-CoT supervision, and that of Qwen3-14B increased from 5.870 0 to 5.964 2. At the expert blind review level, the overall score of Qwen3-32B (PR-CoT) was 7.026 0±0.107 7, higher than 6.416 3±0.288 9 for its baseline. In zero-shot cross-disease testing, the PR-CoT model still showed advantages in semantic evaluation and expert evaluation on both gynecological and heart-system disorder test sets, indicating a certain degree of transferability. ConclusionThe benefits of PR-CoT supervision are mainly reflected in TCM semantic consistency and clinical plausibility, rather than in improved hard matching of structural elements. These findings support understanding syndrome manifestation recognition as a process of generating and expressing latent pathogenesis structures, rather than as a classification task within a traditional fixed label space. By introducing pathogenesis reasoning as an explicit intermediate structure into the modeling process and combining it with a progressive multidimensional evaluation framework, this study provides a methodological pathway for intelligent TCM syndrome differentiation that integrates theoretical alignment, interpretability, and multi-level evaluation.

ObjectiveTo study the chemical constituents of Bidens pilosa and the in vitro antiproliferative activity of some compounds against gastric cancer cells. MethodsThe chemical constituents were isolated and purified by methods such as silica gel column chromatography， preparative thin layer chromatography， medium pressure preparation chromatography， semi-preparative high performance liquid chromatography（HPLC） and recrystallization， their structures were identified on the basis of physicochemical properties， spectral data and circular dichroism spectra. Thiazole blue（MTT） assay was used to determine the in vitro inhibitory activityies of some isolated compounds against human gastric cancer SGC-7901 cells， and molecular docking was used to predict their potential targets. ResultsTwenty-five compounds were isolated from the petroleum ether fraction of B. pilosa and identified as bidpillignan A（1）， 5α， 8α-epidioxy-（22E， 24R）-ergosta-6， 22-diene-3β-ol（2）， 3β-hydroxysitost-5-en-7-one（3）， （20R）-6β-hydroxystigmasta-4， 22-dien-3-one（4）， 3β-hydroxylstigmasta-5， 22-dien-7-one（5）， （22E， 24S）-24-methyl-5α-choleata-7， 22-diene-3β， 5α， 6β-triol（6）， stigamast-5-ene-3β， 7α-diol（7）， stigmast-5， 22-diene-3β， 7α-diol（8）， 7β-hydroxysitosterol（9）， stigmasterol（10）， artabotrol（11）， cosmosoic acid（12）， ent-4（15）-eudesmene-1β， 6α-diol（13）， clovandiol（14）， 1H-indole-3-carboxaldehyde（15）， 6， 7-dimethoxycoumarin（16）， 3'， 4'-dimethoxyquercetin（17）， 8， 8'-bis-（dihydroconiferyl）-diferuloylate（18）， hydroxydihydrobovolide（19）， 2-deacetyl-11β， 13-dihydroxyxanthinin（20）， loliolide（21）， pubescone（22）， （+）-dehydrovomifoliol（23）， 2， 3-dihydroxypropyl palmitate（24） and n-hexadecane acid（25）. Among them， compound 1 was a new compound， compounds 3-9， 11-12， 18-20， 22-23 were first isolated from Bidens genus plants， and compounds 13 and 14 were isolated from this plant for the first time. Compounds 17 and 18 showed good in vitro proliferation inhibitory activities against SGC-7901 cells with the half inhibitory concentrations（IC₅₀） of 3.11， 7.22 μmol·L^-1， respectively. Molecular docking results showed that the potential targets of the two exerting anti-gastric cancer activity might be vascular endothelial growth factor receptor 2（VEGFR2） and poly（adenosine diphosphate-ribose） polymerase 7（PARP7）. ConclusionOne new compound， 14 genus first compounds， and 2 species first compounds were isolated from B. pilosa， among which compounds 17 and 18 processed anti-gastric cancer cell proliferation activity in vitro， and the targets may be VEGFR2 and PARP7.

Colorimetric analysis is a detection and quantification method based on observable color changes in response to analytes,which offers significant advantages including visually detectable signals,straightforward operation,rapid response,and low cost.Consequently,it plays a crucial role in a variety of fields.With increasingly diverse and complex application,colorimetric analysis requires continuous improvement in sensitivity,adaptability to diverse detection environments,and complex data handling capabilities.In recent years,the development of artificial intelligence technology,particularly within its core domain of machine learning(ML),has led to significant advancements in colorimetric analysis.The ML-assisted colorimetric analysis enables high-throughput and high-sensitivity detection,alongside automated analysis,thereby providing novel strategies to overcome the inherent limitations.This review categorized machine learning techniques and summarized their application in colorimetric analysis,introducing two fundamental categories of supervised learning,and unsupervised learning based on the division of core learning paradigms.The research progress of ML-assisted colorimetric analysis in the fields of environmental monitoring,biochemical detection,and food safety were summarized.Finally,the current challenges facing by this research area were analyzed and the research prospect of ML-assisted colorimetric analysis was outlined.

Background: This study explores effective fixation methods for Pauwel type III femoral neck fractures by evaluating the biomechanical benefits of adding a screw to the Femoral Neck System (FNS). Methods: Computed tomography (CT) scans of an 82-year-old female patient with an intertrochanteric fracture were used to establish a finite element femur model with heterogeneous material properties. Finite element models of Pauwel type III fractures were created with and without an additional screw. The central and inferior trajectories of the FNS bolt were examined separately and combined with an additional screw for virtual fixation. Walking and stair-climbing loads were applied. Results: With the addition of a screw, both peak maximum and minimum principal strains consistently stayed comparable or decreased in models with both central and inferior bolt trajectories, while the volume of elements with principal strain exceeding 1% decreased by more than half. The peak von Mises stress observed in the implants ranged from 215.7 to 359.3 MPa, remaining below the titanium alloy's yield strength of 800 MPa. For normal walking, the addition of a screw to the central bolt trajectory model decreased the fracture gap by 50.6% and reduced sliding distance by 8.6%. For the inferior bolt trajectory, the gap was reduced by 57.9% and sliding distance by 25.0%.Under stair-climbing conditions, these improvements were also evident; the central trajectory model saw a halved fracture gap and a 7.9% decrease in sliding distance, while the inferior trajectory model experienced a 55.7% gap reduction and a 27.2% decrease in sliding distance. The additional screw increased the area ratio of the fracture site experiencing interfragmentary compression 34%–39%, while the additional screw alleviated peak interfragmentary compression by 12%–18% under both normal walking and stair-climbing conditions. Conclusions The addition of a screw reduced the fracture gap, sliding distance, and peak interfragmentary compression, while increasing the area ratio of interfragmentary compression under both walking and stair-climbing loads, regardless of the FNS bolt trajectory, suggesting a better mechanical environment for fracture healing.

Background/Aims: Although numerous noninvasive steatosis indices have been developed to assess hepatic steatosis, whether they can be applied to young adults in the evaluation of metabolic dysfunction-associated steatotic liver disease (MASLD) remains uncertain. Methods: Data from patients under 35 years of age who visited the Liver Health Clinic at the Armed Forces Goyang Hospital between July 2022 and January 2024 were retrospectively collected. Steatosis was diagnosed on the basis of a controlled attenuation parameter score ≥250dB/m. MASLD was defined as the presence of steatosis in patients with at least one cardiometabolic risk factor. Results: Among the 1,382 study participants, 901 were diagnosed with MASLD. All eight indices for diagnosing steatosis differed significantly between the MASLD and non-MASLD groups (p<0.001). Regarding the predictive performance, the hepatic steatosis index (HSI), fatty liver index (FLI), Framingham steatosis index, Dallas steatosis index, Zhejiang University index, lipid accumulation product, visceral adiposity index, and triglyceride glucose-body mass index exhibited an area under the curve of 0.898, 0.907, 0.899, 0.893, 0.915, 0.869, 0.791, and 0.898, respectively. The cutoff values for the FLI and HSI were re-examined, indicating a need for alternative cutoff values for the HSI, with a rule-in value of 42 and a rule-out value of 36 in this population. Conclusions This study presents novel findings regarding the predictive performance of established steatosis markers in young adults. Alternative cutoff values for the HSI in this population have been proposed and warrant further validation.

Background/Aims: This study aimed to evaluate the performance of the Model for End-Stage Liver Disease (MELD) 3.0 for predicting mortality and liver-related complications compared with the Child-Pugh classification, albumin-bilirubin (ALBI) grade, the MELD, and the MELD sodium (MELDNa) score. Methods: We evaluated a multicenter retrospective cohort of incorporated patients with cirrhosis between 2013 and 2019. We conducted comparisons of the area under the receiver operating characteristic curve (AUROC) of the MELD3.0 and other models for predicting 3-month mortality. Additionally, we assessed the risk of cirrhosis-related complications according to the MELD3.0 score. Results: A total of 3,314 patients were included. The mean age was 55.9±11.3 years, and 70.2% of the patients were male. Within the initial 3 months, 220 patients (6.6%) died, and the MELD3.0had the best predictive performance among the tested models, with an AUROC of 0.851, outperforming the Child-Pugh classification, ALBI grade, MELD, and MELDNa. A high MELD3.0score was associated with an increased risk of mortality. Compared with that of the group with a MELD3.0 score <10 points, the adjusted hazard ratio of the group with a score of 10–20 pointswas 2.176, and that for the group with a score of ≥20 points was 4.892. Each 1-point increase inthe MELD3.0 score increased the risk of cirrhosis-related complications by 1.033-fold. The risk of hepatorenal syndrome showed the highest increase, with an adjusted hazard ratio of 1.149, followed by hepatic encephalopathy and ascites. Conclusions The MELD3.0 demonstrated robust prognostic performance in predicting mortality in patients with cirrhosis. Moreover, the MELD3.0 score was linked to cirrhosis-related complications, particularly those involving kidney function, such as hepatorenal syndrome and ascites.