1.Explainable Machine Learning Model for Predicting Prognosis in Patients with Malignant Tumors Complicated by Acute Respiratory Failure: Based on the eICU Collaborative Research Database in the United States
Zihan NAN ; Linan HAN ; Suwei LI ; Ziyi ZHU ; Qinqin ZHU ; Yan DUAN ; Xiaoting WANG ; Lixia LIU
Medical Journal of Peking Union Medical College Hospital 2026;17(1):98-108
To develop and validate a model for predicting intensive care unit (ICU) mortality risk in patients with malignant tumors complicated by acute respiratory failure (ARF) based on an explainable machine learning framework. Clinical data of patients with malignant tumors and ARF were extracted from the eICU Collaborative Research Database in the United States, including demographic characteristics, comorbidities, vital signs, laboratory test indicators, and major interventions within the first 24 hours after ICU admission.The study outcome was ICU death.Enrolled patients were randomly divided into a training set and a validation set at a ratio of 7:3.Predictor variables were selected using least absolute shrinkage and selection operator (LASSO) regression.Five machine learning algorithms-extreme gradient boosting (XGBoost), support vector machine (SVM), Logistic regression, multilayer perceptron (MLP), and C5.0 Decision Tree-were employed to construct predictive models.Model performance was evaluated based on the area under the receiver operating characteristic curve (AUC), accuracy, sensitivity, and other metrics.The optimal model was further interpreted using the Shapley additive explanations (SHAP) algorithm. A total of 3196 patients with malignant tumors complicated by ARF were included.The training set comprised 2, 261 patients and the validation set 935 patients; 683 patients died during ICU stay, while 2513 survived.LASSO regression ultimately selected 12 variables closely associated with patient ICU outcomes, including sepsis comorbidity, use of vasoactive drugs, and within the first 24 hours after ICU admission: minimum mean arterial pressure, maximum heart rate, maximum respiratory rate, minimum oxygen saturation, minimum serum bicarbonate, minimum blood urea nitrogen, maximum white blood cell count, maximum mean corpuscular volume, maximum serum potassium, and maximum blood glucose.After model evaluation, the XGBoost model demonstrated the best performance.The AUCs for predicting ICU mortality risk in the training and validation sets were 0.940 and 0.763, respectively; accuracy was 88.3% and 81.2%;sensitivity was 98.5% and 95.9%.Its predictive performance also remained optimal in sensitivity analyses.SHAP analysis indicated that the top five variables contributing to the model's predictions were minimum oxygen saturation, minimum serum bicarbonate, minimum mean arterial pressure, use of vasoactive drugs, and maximum white blood cell count. This study successfully developed a mortality risk prediction model for ICU patients with malignant tumors complicated by ARF based on a large-scale dataset and performed explainability analysis.The model aids clinicians in early identification of high-risk patients and implementing individualized interventions.
2.Dual activation of GCGR/GLP1R signaling ameliorates intestinal fibrosis via metabolic regulation of histone H3K9 lactylation in epithelial cells.
Han LIU ; Yujie HONG ; Hui CHEN ; Xianggui WANG ; Jiale DONG ; Xiaoqian LI ; Zihan SHI ; Qian ZHAO ; Longyuan ZHOU ; JiaXin WANG ; Qiuling ZENG ; Qinglin TANG ; Qi LIU ; Florian RIEDER ; Baili CHEN ; Minhu CHEN ; Rui WANG ; Yao ZHANG ; Ren MAO ; Xianxing JIANG
Acta Pharmaceutica Sinica B 2025;15(1):278-295
Intestinal fibrosis is a significant clinical challenge in inflammatory bowel diseases, but no effective anti-fibrotic therapy is currently available. Glucagon receptor (GCGR) and glucagon-like peptide 1 receptor (GLP1R) are both peptide hormone receptors involved in energy metabolism of epithelial cells. However, their role in intestinal fibrosis and the underlying mechanisms remain largely unexplored. Herein GCGR and GLP1R were found to be reduced in the stenotic ileum of patients with Crohn's disease as well as in the fibrotic colon of mice with chronic colitis. The downregulation of GCGR and GLP1R led to the accumulation of the metabolic byproduct lactate, resulting in histone H3K9 lactylation and exacerbated intestinal fibrosis through epithelial-to-mesenchymal transition (EMT). Dual activating GCGR and GLP1R by peptide 1907B reduced the H3K9 lactylation in epithelial cells and ameliorated intestinal fibrosis in vivo. We uncovered the role of GCGR/GLP1R in regulating EMT involved in intestinal fibrosis via histone lactylation. Simultaneously activating GCGR/GLP1R with the novel dual agonist peptide 1907B holds promise as a treatment strategy for alleviating intestinal fibrosis.
3.Therapeutic Effect of Cranial Painkiller Pills' Extract Powder in Treatment of Trigeminal Neuralgia Induced by Injection of Talci Pulvis into Infraorbital Foramen of Model Rats Based on OTULIN-regulated Neuroinflammation
Shuran LI ; Xinwei WANG ; Jing SUN ; Dan XIE ; Ronghua ZHAO ; Lei BAO ; Zihan GENG ; Qiyue SUN ; Jingsheng ZHANG ; Yaxin WANG ; Xihe CUI ; Xinying LI ; Bing HAN ; Tianjiao LU ; Xiaolan CUI ; Liying LIU ; Shanshan GUO
Chinese Journal of Experimental Traditional Medical Formulae 2025;31(21):21-28
ObjectiveThis paper aims to verify the therapeutic effect of Cranial Painkiller pills' extract powder prepared by the new process on the rat's trigeminal neuralgia model caused by infraorbital injection of Talci Pulvis, evaluate its potential clinical application value, and compare the therapeutic effect with that of Cranial Painkiller granules, so as to provide data support for the application of the Cranial Painkiller pills' extract powder and precise treatment. MethodsThe rat's trigeminal neuralgia model was constructed by infraorbital injection of Talci Pulvis, and the rats were randomly divided into the normal group, model group, carbamazepine group (60 mg·kg-1), Cranial Painkiller granules group (2.70 g·kg-1), and low, medium, and high dosage groups of Cranial Painkiller pills' extract powder (1.35, 2.70, 5.40 g·kg-1) according to the basal mechanical pain thresholds, and there were 10 rats in each group. The drug was administered by gavage to each group 2 h after modeling, and distilled water was given by gavage to the normal and model groups under the same conditions once a day for 10 d. Von Frey brushes were used to measure mechanical pain thresholds in rats. Hematoxylin-eosin (HE) staining was used to detect pathological changes in the trigeminal ganglion, and enzyme-linked immunosorbent assay (ELISA) was used to detect the inflammatory factors interleukin-1 (IL-1), interleukin-6 (IL-6), interleukin-8 (IL-8), and tumor necrosis factor-α (TNF-α) levels in rat serum, as well as neuropeptide substance P (SP) and β-endorphin (β-EP) levels in rat brain tissue. Western blot technique was used to detect the levels of NLRP3, ASC, Caspase-1, and OTULIN proteins in rat brain tissue. ResultsCompared with the normal group, the pain threshold of rats in the model group showed a continuous significant decrease (P<0.01). The pathological damage of brain tissue was significant (P<0.01), and the inflammatory levels of IL-1, IL-6, IL-8, and TNF-α in serum were significantly elevated (P<0.01). The level of the SP in the brain tissue was significantly elevated (P<0.01), and the level of β-EP was significantly reduced (P<0.01), while the level of OTULIN was significantly reduced, and NLRP3, ASC, and Caspase-1 protein levels were significantly elevated (P<0.01). After administration of the drug, compared with the model group, the pain threshold of each dose group of the Cranial Painkiller pills' extract powder and the Cranial Painkiller granules group significantly increased (P<0.01). The inflammatory levels of IL-1, IL-6, IL-8, and TNF-α and SP levels significantly decreased (P<0.01), and the β-EP levels were significantly elevated (P<0.01), while the levels of OTULIN protein were significantly elevated (P<0.05, P<0.01), and the levels of NLRP3, ASC proteins were decreased (P<0.01)in high dose Cranial Painkiller pills' extract powder. Meanwhile, compared with those in the model group, the trigeminal ganglion lesions of rats in the Cranial Painkiller pills' extract powder and Cranial Painkiller granules groups showed different degrees of improvement (P<0.05, P<0.01). ConclusionThe Cranial Painkiller pills' extract powder has significant therapeutic effects on the rat model of trigeminal neuralgia induced by infraorbital injection of Talci Pulvis, and its mechanism is related to the improvement of OTULIN-regulated neuroinflammation.
4.Therapeutic Effect of Cranial Painkiller Pills' Extract Powder in Treatment of Trigeminal Neuralgia Induced by Injection of Talci Pulvis into Infraorbital Foramen of Model Rats Based on OTULIN-regulated Neuroinflammation
Shuran LI ; Xinwei WANG ; Jing SUN ; Dan XIE ; Ronghua ZHAO ; Lei BAO ; Zihan GENG ; Qiyue SUN ; Jingsheng ZHANG ; Yaxin WANG ; Xihe CUI ; Xinying LI ; Bing HAN ; Tianjiao LU ; Xiaolan CUI ; Liying LIU ; Shanshan GUO
Chinese Journal of Experimental Traditional Medical Formulae 2025;31(21):21-28
ObjectiveThis paper aims to verify the therapeutic effect of Cranial Painkiller pills' extract powder prepared by the new process on the rat's trigeminal neuralgia model caused by infraorbital injection of Talci Pulvis, evaluate its potential clinical application value, and compare the therapeutic effect with that of Cranial Painkiller granules, so as to provide data support for the application of the Cranial Painkiller pills' extract powder and precise treatment. MethodsThe rat's trigeminal neuralgia model was constructed by infraorbital injection of Talci Pulvis, and the rats were randomly divided into the normal group, model group, carbamazepine group (60 mg·kg-1), Cranial Painkiller granules group (2.70 g·kg-1), and low, medium, and high dosage groups of Cranial Painkiller pills' extract powder (1.35, 2.70, 5.40 g·kg-1) according to the basal mechanical pain thresholds, and there were 10 rats in each group. The drug was administered by gavage to each group 2 h after modeling, and distilled water was given by gavage to the normal and model groups under the same conditions once a day for 10 d. Von Frey brushes were used to measure mechanical pain thresholds in rats. Hematoxylin-eosin (HE) staining was used to detect pathological changes in the trigeminal ganglion, and enzyme-linked immunosorbent assay (ELISA) was used to detect the inflammatory factors interleukin-1 (IL-1), interleukin-6 (IL-6), interleukin-8 (IL-8), and tumor necrosis factor-α (TNF-α) levels in rat serum, as well as neuropeptide substance P (SP) and β-endorphin (β-EP) levels in rat brain tissue. Western blot technique was used to detect the levels of NLRP3, ASC, Caspase-1, and OTULIN proteins in rat brain tissue. ResultsCompared with the normal group, the pain threshold of rats in the model group showed a continuous significant decrease (P<0.01). The pathological damage of brain tissue was significant (P<0.01), and the inflammatory levels of IL-1, IL-6, IL-8, and TNF-α in serum were significantly elevated (P<0.01). The level of the SP in the brain tissue was significantly elevated (P<0.01), and the level of β-EP was significantly reduced (P<0.01), while the level of OTULIN was significantly reduced, and NLRP3, ASC, and Caspase-1 protein levels were significantly elevated (P<0.01). After administration of the drug, compared with the model group, the pain threshold of each dose group of the Cranial Painkiller pills' extract powder and the Cranial Painkiller granules group significantly increased (P<0.01). The inflammatory levels of IL-1, IL-6, IL-8, and TNF-α and SP levels significantly decreased (P<0.01), and the β-EP levels were significantly elevated (P<0.01), while the levels of OTULIN protein were significantly elevated (P<0.05, P<0.01), and the levels of NLRP3, ASC proteins were decreased (P<0.01)in high dose Cranial Painkiller pills' extract powder. Meanwhile, compared with those in the model group, the trigeminal ganglion lesions of rats in the Cranial Painkiller pills' extract powder and Cranial Painkiller granules groups showed different degrees of improvement (P<0.05, P<0.01). ConclusionThe Cranial Painkiller pills' extract powder has significant therapeutic effects on the rat model of trigeminal neuralgia induced by infraorbital injection of Talci Pulvis, and its mechanism is related to the improvement of OTULIN-regulated neuroinflammation.
5.Analysis of the demands for pharmaceutical clinic service and influential factors based on Kano model
Han SHAN ; Xuan YE ; Zihan GUO ; Jing WU ; Jinwei HU ; Xiaopei TONG ; Yufei BIN ; Jiyong LIU ; Qiong DU ; Mengmeng WANG
China Pharmacy 2025;36(22):2850-2855
OBJECTIVE To explore the characteristics and influential factors of pharmaceutical clinic service demands, providing evidence for optimizing pharmaceutical service models and facilitating pharmaceutical service models of pharmacist role transformation. METHODS A cross-sectional survey design was adopted, and 410 outpatient participants were selected from Fudan University Shanghai Cancer Center through convenience sampling for questionnaire administration from February to May 2025. Kano model was applied to analyze the demand attributes of 25 pharmaceutical services, while questionnaires were used to assess patients’ awareness and demand status. Subgroup analyses were conducted based on key demographic variables such as gender, age, educational attainment, and economic burdens, to SACA- systematically examine the differences in Kano attribute classification among patients in each subgroup. RESULTS The awareness rate of pharmaceutical outpatient services among patients was only 14.63%, yet those who were aware demonstrated a significantly higher demand rate for such services compared to those who were unaware (P<0.001). The demand for pharmaceutical clinic services exhibited a hierarchical characteristic: twelve items were identified as attractive attributes (e. g., providing suggestions for more affordable treatment options, offering online consultation services, etc.), five items as expected attributes (e.g., having a good attitude and being able to patiently answer your questions, etc.), three items as must-have attributes (e.g., providing guidance on medication dosage and usage, providing guidance on medication precautions, etc.), five items as indifferent attributes (e.g., providing treatment plan recommendations based on the patient’s condition). There were zero items classified as reverse attribute. Subgroup analysis revealed that female patients showed greater concern for “neat and clean attire of medical staff” than male patients (P<0.001); patients under 60 years of age demonstrated stronger demand for “providing treatment plan recommendations based on patients’ conditions” compared to patients aged 60 or above (P=0.016); those with below high school education placed greater emphasis on “providing guidance on medication precautions” compared to those with a high school education or above (P=0.011); patients with lower economic burdens exhibited stronger preferences for “neat and clean attire of medical staff ” (P=0.002). CONCLUSIONS The public awareness rate of pharmaceutical clinic services is considerably low; however, those who are aware of such services demonstrate significantly higher demand. The medication safety-related services and convenience-oriented demands should be prioritized in the development of pharmaceutical clinics. Moreover, the study also revealed that factors such as gender, age, educational level, and economic burdens exert significant influences on patients’ service demands.
6.GPT2-ICC: A data-driven approach for accurate ion channel identification using pre-trained large language models.
Zihan ZHOU ; Yang YU ; Chengji YANG ; Leyan CAO ; Shaoying ZHANG ; Junnan LI ; Yingnan ZHANG ; Huayun HAN ; Guoliang SHI ; Qiansen ZHANG ; Juwen SHEN ; Huaiyu YANG
Journal of Pharmaceutical Analysis 2025;15(8):101302-101302
Current experimental and computational methods have limitations in accurately and efficiently classifying ion channels within vast protein spaces. Here we have developed a deep learning algorithm, GPT2 Ion Channel Classifier (GPT2-ICC), which effectively distinguishing ion channels from a test set containing approximately 239 times more non-ion-channel proteins. GPT2-ICC integrates representation learning with a large language model (LLM)-based classifier, enabling highly accurate identification of potential ion channels. Several potential ion channels were predicated from the unannotated human proteome, further demonstrating GPT2-ICC's generalization ability. This study marks a significant advancement in artificial-intelligence-driven ion channel research, highlighting the adaptability and effectiveness of combining representation learning with LLMs to address the challenges of imbalanced protein sequence data. Moreover, it provides a valuable computational tool for uncovering previously uncharacterized ion channels.
7.Ferroptosis Mechanism of Knee Osteoarthritis and Its Progress in Traditional Chinese Medicine Intervention
Feiyu HAN ; Yujing CAO ; Zihan LIN ; Yinlong WANG ; Wuyue TONG
Journal of Guangzhou University of Traditional Chinese Medicine 2025;42(5):1290-1296
Knee osteoarthritis(KOA)is a common musculoskeletal disease,and ferroptosis is a new type of programmed cell death involving iron ions.This paper summarized the relationship between the pathological changes of ferroptosis and the onset of KOA,and reviewed the progress of traditional Chinese medicine(TCM)intervention in KOA based on the pathological changes of ferroptosis.The pathological changes of ferroptosis are mainly related to the imbalance of amino acid antioxidant system,iron metabolism disorder and lipid peroxide accumulation.The cystine/glutamate antiporter(System XC-),glutathione(GSH)and glutathione peroxidase 4(GPX4)are the core targets for regulating ferroptosis,involving the main signaling pathways for studying the TCM intervention of KOA via ferroptosis.Single herbs such as Drynariae Rhizoma,Scutellariae Barbatae Herba,Epimedii Folium,Astragali Radix,Ligustri Lucidi Fructus,and Achyranthis Bidentatae Radix,and multi-herb formulae such as Duhuo Jisheng Decoction and Yougui Pills,can inhibit the ferroptosis of chondrocytes through multiple targets and multiple pathways,thus treating KOA.The research of the TCM intervention on the ferroptosis of chondrocytes and its associated signaling pathways can provides a novel approach for expanding the TCM therapy of KOA.
8.GPT2-ICC:A data-driven approach for accurate ion channel identification using pre-trained large language models
Zihan ZHOU ; Yang YU ; Chengji YANG ; Leyan CAO ; Shaoying ZHANG ; Junnan LI ; Yingnan ZHANG ; Huayun HAN ; Guoliang SHI ; Qiansen ZHANG ; Juwen SHEN ; Huaiyu YANG
Journal of Pharmaceutical Analysis 2025;15(8):1800-1809
Current experimental and computational methods have limitations in accurately and efficiently classi-fying ion channels within vast protein spaces.Here we have developed a deep learning algorithm,GPT2 Ion Channel Classifier(GPT2-ICC),which effectively distinguishing ion channels from a test set con-taining approximately 239 times more non-ion-channel proteins.GPT2-ICC integrates representation learning with a large language model(LLM)-based classifier,enabling highly accurate identification of potential ion channels.Several potential ion channels were predicated from the unannotated human proteome,further demonstrating GPT2-ICC's generalization ability.This study marks a significant advancement in artificial-intelligence-driven ion channel research,highlighting the adaptability and effectiveness of combining representation learning with LLMs to address the challenges of imbalanced protein sequence data.Moreover,it provides a valuable computational tool for uncovering previously uncharacterized ion channels.
9.Application of blepharoplasty by "double-line method" in improving upper eyelid skin relaxation
Zihan XU ; Shengzhi MU ; Feng HAN ; Guodong WANG ; Jianwu LI ; Liang WANG ; Na WANG
Chinese Journal of Plastic Surgery 2025;41(1):15-19
Objective:To investigate the efficacy of blepharoplasty using the "double-line technique" in addressing upper eyelid skin laxity in middle-aged and elderly patients.Methods:A retrospective analysis was conducted on the clinical data of middle-aged and elderly patients with moderate to severe upper eyelid skin laxity who were admitted to the Plastic Surgery Department of Shaanxi Provincial People’s Hospital between January 2020 and January 2023. The same surgeon performed the "double-line" blepharoplasty to excise excess upper eyelid skin for laxity correction. Patients were positioned flat with their eyes closed; an arcuate line a (5-7 mm above the pupil) and a second line b (aligned with the eyelid margin in an upright position) were marked. The skin between these two lines was excised. Postoperatively, the double eyelid shape, improvement in skin laxity, adverse reactions, and patient satisfaction (categorized as satisfied, basically satisfied, unsatisfied) were assessed.Results:A total of 145 patients were included in the study, comprising 27 males and 118 females. Their ages ranged from 38 to 78 years, with an average age of 52 years. All patients’ incisions healed primarily without complications such as hematoma and infection. During the 3 to 12 months of follow-up after surgery, the double eyelid arcs on both sides appeared symmetrical and natural. The upper eyelid laxity was effectively corrected, obscured vision was improved, crow’s feet were reduced, and the periorbital area was rejuvenated. Patient satisfaction with the surgical results was high, with 112 patients reporting satisfaction, 25 reporting basically satisfied, and 8 reporting dissatisfaction. The overall satisfaction rate reached 94.48%(137/145). Among those dissatisfied, 6 cases cited the double eyelid crease as being too narrow, and 2 cases noted asymmetry in the double eyelid crease.Conclusion:The "double-line method" for blepharoplasty is straightforward to perform and effectively removes excess upper eyelid skin. It is suitable for middle-aged and elderly patients with moderate to severe upper eyelid skin laxity, resulting in favorable surgical outcomes.
10.Research on Service Mode of Public Hospital Operation Assistant Working Team Based on S-MDT
Zihan MU ; Zhiping GUO ; Yuxiu TAO ; Leichao WANG ; Xueqing YU ; Zixu GUO ; Han LIU ; Yaojun ZHAO
Chinese Hospital Management 2025;45(3):32-35
The operation assistant work team of public hospitals is an important bridge and link to realize the integration of industry and finance,and an important practitioner to promote the realization of lean operation management.It examines the service model of the operation assistant team engaged in Similar Multi-disciplinary Treatment,focusing on organizational structure,job responsibilities,and service mechanisms.Build a four-tiered lean operational management system and set up an operation assistant work team that combines production,university and research.Based on specific management matters,establish an operation assistant service mechanism such as problem assessment and classification management mechanism,project standardization management mechanism,reward and punishment assessment mechanism,etc.,to promote the optimization and integration of hospital medical education,research,prevention and management and core elements such as human,talent,material and technology resources.Help public hospitals to continuously improve their lean operation management level.

Result Analysis
Print
Save
E-mail