Identifying drug-drug interactions (DDIs) is essential to prevent adverse effects from polypharmacy. Although deep learning has advanced DDI identification, the gap between powerful models and their lack of clinical application and evaluation has hindered clinical benefits. Here, we developed a Multi-Dimensional Feature Fusion model named MDFF, which integrates one-dimensional simplified molecular input line entry system sequence features, two-dimensional molecular graph features, and three-dimensional geometric features to enhance drug representations for predicting DDIs. MDFF was trained and validated on two DDI datasets, evaluated across three distinct scenarios, and compared with advanced DDI prediction models using accuracy, precision, recall, area under the curve, and F1 score metrics. MDFF achieved state-of-the-art performance across all metrics. Ablation experiments showed that integrating multi-dimensional drug features yielded the best results. More importantly, we obtained adverse drug reaction reports uploaded by Xiangya Hospital of Central South University from 2021 to 2023 and used MDFF to identify potential adverse DDIs. Among 12 real-world adverse drug reaction reports, the predictions of 9 reports were supported by relevant evidence. Additionally, MDFF demonstrated the ability to explain adverse DDI mechanisms, providing insights into the mechanisms behind one specific report and highlighting its potential to assist practitioners in improving medical practice.

Objective:To explore the high-risk factors for parenteral nutrition associated cholestasis（PNAC）in extremely/ultra-low birth weight infants，and establish a risk Alignment Diagram prediction model.Methods:We retrospectivly analyzed the clinical data of hospitalized extremely/ultra-low birth weight infants admitted to Neonatology Department at Quanzhou Children's Hospital from January 2019 to December 2020，using multivariate Logistic regression analysis to screen for independent risk factors for the occurrence of PNAC.An Alignment Diagram model prediction model for PNAC was constructed by using R software，and the performance of the model was evaluated through receiver operating characteristic curves.Results:A total of 203 extremely/ultra-low birth weight infants were included，with a median gestational age of 29.14（28.00，30.86）weeks and a median birth weight of 1　170（1　000,1　300）g.Among them，26（12.81%）cases developed PNAC.Multivariate Logistic regression analysis showed that the duration of parenteral nutrition（ OR=1.015 ，95% CI 1.003-1.034），the cumulative amount of glucose（ OR=1.014 ，95% CI 1.001-1.028），small for gestational age（ OR=3.455 ，95% CI 1.127-10.589），and neonatal sepsis（ OR=3.142 ，95% CI 1.039-9.503）were independent risk factors for PNAC（ P＜0.05）；The four independent risk factors mentioned above were introduced into R software to construct an Alignment Diagram model，the area under the receiver operating characteristic curve was 0.835（95% CI 0.842-0.731），and the results of the Hosmer Limeshow goodness of fit test show that：χ 2=5.34，degree of freedom=8， P=0.72.A calibration curve indicated good consistency between the predicted probability of the model and the actual occurrence rate，with good accuracy. Conclusion:The Alignment Diagram model constructed based on four independent risk factors of the duration of parenteral nutrition，glucose accumulation，small for gestational age infants，and neonatal sepsis exhibits high predictive ability，and is expected to provide an intuitive and convenient visualization tool for preventing or reducing the occurrence of PNAC in extremely/ultra-low birth weight infants

Objective To investigate the involvement of calmodulin(CaM)in the pathogenesis of Alzheimer disease(AD)and the mechanism by which CaM binds to amyloid-β(Aβ).Methods The hub genes expressed in AD and predicted to be the target proteins for AD prevention and treatment were obtained using bioinformatics methods.The GST-Aβ1-42 recombinant plasmid was constructed through genetic recombination and was then sequenced.The recombinant plasmids were identified using agarose gel electrophoresis,while the extracted and purified GST-Aβ1-42 fusion protein was confirmed using SDS-PAGE gel electrophoresis.GST pull-down assay was used to detect the interaction between GST-Aβ1-42 protein and CaM,expressed in the plasmid.Results The top 20 hub genes in degree ranking were obtained.The DNA sequencing results of the plasmid proved that the recombinant plasmid was successfully constructed.The agarose gel electrophoresis results indicated that the fragment digested by the enzyme was similar to the molecular weight of the Aβ1-42 gene seg-ments,further proving the successful construction of the recombinant plasmid.Binding of GST-Aβ1-42 protein to CaM in a concentration dependent manner was revealed through the GST pull down experiment.Conclusion The GST-Aβ1-42 recombinant plasmid is success-fully constructed and is shown to bind to CaM.

Objective:To investigate the influencing factors for lymph node metastasis and prognosis in early gastric cancer.Methods:The retrospective cohort study was conducted. The clinicopathological data of 1 011 patients with early gastric cancer who were admitted to the Zhejiang Cancer Hospital from January 2010 to December 2019 were collected. There were 561 males and 450 females, aged (58±11)years. All patients underwent radical resection of gastric cancer and the lymph node metastasis of each group was identified according to the pathological examination on patients' surgical specimens. Observation indicators: (1) lymph node metastasis in early gastric cancer; (2) influencing factors for lymph node metastasis in early gastric cancer; (3) influencing factors for prognosis in early gastric cancer. Measurement data with normal distribution were represented as Mean± SD, and comparison between groups was conducted using the t test. Count data were described as absolute numbers or percentages, and comparison between groups was conducted using the chi-square test. Comparison of ordinal data was analyzed using the non‐parameter rank sum test. Univariate analysis was conducted using the Log-Rank test and Logistic regression model, and multivariate analysis was conducted using the Logistic stepwise regression model and COX step-wise regression model. The Kaplan-Meier method was used to calculate survival rate and draw survival curve, and the Log-Rank test was used for survival analysis. Results:(1) Lymph node metastasis in early gastric cancer. The lymph node metastasis rate of 1 011 patients with early gastric cancer was 23.640%(239/1 011), in which the lymph node metastasis rate of patients with T1a stage gastric cancer was 11.883%(53/446), and the lymph node metastasis rate of patients with T1b stage gastric cancer was 32.920%(186/565). There were 239 patients with lymph node metastasis mainly concentrated in the first station, including 7 cases with No.1 lymph node metastasis, 11 cases with No.2 lymph node metastasis, 135 cases with No.3 lymph node metastasis, 59 cases with No.4 lymph node metastasis, 39 cases with No.5 lymph node metastasis, 91 cases with No.6 lymph node metastasis, 6 cases with No.7 lymph node metastasis, 8 cases with No.8 lymph node metastasis, 8 cases with No.9 lymph node metastasis and 6 cases with No.10 lymph node metastasis. Multiple lymph node metastases may exist in the same patient. For lymph node metastasis in different tumor sites, there were 4 cases, 2 cases and 1 case of lymph node metastasis in the No.2, 3 and 5 lymph node in patients with upper gastric cancer. There were 3 cases, 7 cases, 36 cases, 15 cases, 3 cases and 5 cases of lymph node metastasis in the No.1, 2, 3, 4, 5 and 6 lymph node in patients with middle gastric cancer. There were 4 cases, 97 cases, 44 cases, 35 cases and 86 cases of lymph node metastasis in the No.1, 3, 4, 5 and 6 lymph node in patients with lower gastric cancer. (2) Influencing factors for lymph node metastasis in early gastric cancer. Results of multivariate analysis showed that tumor diameter, tumor location, degree of tumor invasion, vascular thrombus, degree of tumor differentiation were independent factors influencing lymph node metastasis in early gastric cancer ( odds ratio=1.80, 1.49, 2.65, 5.76, 0.60, 95% confidence interval as 1.29-2.50, 1.11-2.00, 1.81-3.88, 3.87-8.59, 0.48-0.76, P<0.05). (3) Influencing factors for prognosis in early gastric cancer. All 1 011 patients were followed up for 43(range, 0-135)months, and the 3-year overall survival rate was 97.32%. Results of multivariate analysis showed that age >60 years and lymph node metastasis were independent risk factors influencing prognosis in early gastric cancer ( hazard ratio=9.50, 2.20, 95% confidence interval as 3.31-27.29, 1.00-4.87, P<0.05). Results of further analysis showed that the 3-year overall survival rate was 99.37% and 94.66% in patient with age >60 years and ≤60 years, respectively, showing a significant difference between them ( χ2=25.33, P<0.05). The 3-year overall survival rate was 95.42% and 97.92% in patients with and without lymph node metastasis, respectively, showing a significant difference between them ( χ2=5.69, P<0.05). Conclusions:The lymph node metastasis rate of early gastric cancer can reach 23.640%. Tumor diameter, tumor location, degree of tumor invasion, vascular thrombus, degree of tumor differentia-tion are independent factors influencing lymph node metastasis in early gastric cancer, age >60 years and lymph node metastasis are independent risk factors influencing prognosis.

ObjectiveTo investigate the effect of intensive insulin therapy in critically ill elderly patients.MethodsElderly patients ( ≥ 65 years) admitted to the ICU of Beijing Tongren Hospital from June 2005 to December 2007 were divided into Group A ( glucose control target was 4.4-6.1mmol/L) and Group B ( glucose control target was 7.3-8.3mmol/L).Blood glucose level was controlled with a computer-assisted glucose control protocoL ResultsA total of 639 patients were enrolled,of which 280 were in Group A and 359 in Group B.The mean blood glucose level of the 2 groups was (6.07 ± 0.56) mmol/L and (7.52 ± 0.87 ) mmol/L respectively,both within the target ranges.The hyperglycemic index was (0.69±0.44) mmol/L in Group A and ( 1.60 ±0.73) mmol/L in Group B (P =0.000).No hypoglycemia adverse events occurred in either group.No significant differences were observed in the length of stay in ICU,duration of mechanical ventilation,hospitalization expenses,ICU mortality,and hospital mortality of the 2 groups.ConclusionMaintaining the blood glucose level of critically ill elderly patients at ≤8.3 mmol/L is safe and practical.