Objective:To preliminarily establish a sample clot warning model for coagulation screening tests using 5 machine learning methods.Methods:This cross-sectional study collected 7 401 routine screening test samples from Peking Union Medical College Hospital, Chinese Academy of Medical Sciences, from January 1st, 2015, to August 18th, 2024, including 4 786 clotted (positive) and 2 615 qualified (negative) samples for model development. The dataset was divided into Dataset 1 and Dataset 2 based on a reagent change for APTT in December 2018, with separate models developed for each. An additional 2 493 samples (October 31st to November 8th, 2024) were used to evaluate consistency between the model and manual assessment, while 23 200 samples (October 17th to December 31st, 2024) were used for assessing real-world predictive performance. Five machine learning algorithms were employed to develop the clot prediction model: logistic regression (LR), random forest (RF), extreme gradient boosting (XGBoost), naive bayes (NB), and artificial neural network (ANN), with the ANN model constructed using two different hidden layer and neuron parameter settings. Model selection was based on AUC, accuracy, sensitivity, specificity, F1-score, PPV, and NPV, with the optimal model integrated into the LIS for validation.Results:Among the six models using 5 machine learning algorithms, XGBoost demonstrated the highest performance (AUC=0.961, sensitivity=0.945, F1-score=0.934) and robustness to reagent changes ( Z=-1.333, P=0.113). When deployed, the differences between the model's predictions and manual pre-judgment were statistically significant ( Z=-5.289 to 8.933, all P<0.01). The predictive efficacy indices AUC (95% CI), sensitivity, specificity, and accuracy of the XGBoost model deployed in real-world operation of the LIS were 0.939 (0.918—0.960), 0.958, 0.921, and 0.921 respectively. Conclusion:In this study, a clot warning model for coagulation screening samples was established based on the XGBoost algorithm, and its prediction efficacy is good, providing a foundation for intelligent pre-analytical quality control for coagulation screening tests.

Objective:To preliminarily establish a sample clot warning model for coagulation screening tests using 5 machine learning methods.Methods:This cross-sectional study collected 7 401 routine screening test samples from Peking Union Medical College Hospital, Chinese Academy of Medical Sciences, from January 1st, 2015, to August 18th, 2024, including 4 786 clotted (positive) and 2 615 qualified (negative) samples for model development. The dataset was divided into Dataset 1 and Dataset 2 based on a reagent change for APTT in December 2018, with separate models developed for each. An additional 2 493 samples (October 31st to November 8th, 2024) were used to evaluate consistency between the model and manual assessment, while 23 200 samples (October 17th to December 31st, 2024) were used for assessing real-world predictive performance. Five machine learning algorithms were employed to develop the clot prediction model: logistic regression (LR), random forest (RF), extreme gradient boosting (XGBoost), naive bayes (NB), and artificial neural network (ANN), with the ANN model constructed using two different hidden layer and neuron parameter settings. Model selection was based on AUC, accuracy, sensitivity, specificity, F1-score, PPV, and NPV, with the optimal model integrated into the LIS for validation.Results:Among the six models using 5 machine learning algorithms, XGBoost demonstrated the highest performance (AUC=0.961, sensitivity=0.945, F1-score=0.934) and robustness to reagent changes ( Z=-1.333, P=0.113). When deployed, the differences between the model's predictions and manual pre-judgment were statistically significant ( Z=-5.289 to 8.933, all P<0.01). The predictive efficacy indices AUC (95% CI), sensitivity, specificity, and accuracy of the XGBoost model deployed in real-world operation of the LIS were 0.939 (0.918—0.960), 0.958, 0.921, and 0.921 respectively. Conclusion:In this study, a clot warning model for coagulation screening samples was established based on the XGBoost algorithm, and its prediction efficacy is good, providing a foundation for intelligent pre-analytical quality control for coagulation screening tests.

Progressive functional deterioration in the cochlea is associated with age-related hearing loss (ARHL). However, the cellular and molecular basis underlying cochlear aging remains largely unknown. Here, we established a dynamic single-cell transcriptomic landscape of mouse cochlear aging, in which we characterized aging-associated transcriptomic changes in 27 different cochlear cell types across five different time points. Overall, our analysis pinpoints loss of proteostasis and elevated apoptosis as the hallmark features of cochlear aging, highlights unexpected age-related transcriptional fluctuations in intermediate cells localized in the stria vascularis (SV) and demonstrates that upregulation of endoplasmic reticulum (ER) chaperon protein HSP90AA1 mitigates ER stress-induced damages associated with aging. Our work suggests that targeting unfolded protein response pathways may help alleviate aging-related SV atrophy and hence delay the progression of ARHL.
Mice ; Animals ; Transcriptome ; Aging/metabolism* ; Cochlea ; Stria Vascularis ; Presbycusis

Animals ; Antlers ; Exosomes ; Mesenchymal Stem Cells ; Osteoarthritis/therapy*

Cell Differentiation ; Humans ; Hyperthermia, Induced ; Stem Cells

Objective To obtain a novel tool-enzyme for genetic engineering with good solubility,strong specificity of enzyme digestion and maintaining the enzyme activity at low temperature by using E.coli expression system to express self-processed re-combinant MBP-HRV 3C fusion protease.Methods The cDNA encoding HRV 3C protease was cloned into pRSF-Duet vector.The recombinant plasmid was transferred into E.coli BL21 (DE3)for expression.HRV 3C protease was obtained through Nichol col-umn affinity purification.The cleavage activity of HRV 3C protease was determined by in vivo experiment.Results HRV 3C prote-ase was highly expressed in E.coli expression system,and the obtained HRV 3C protease could recognize and digest HRV 3C site. Conclusion A novel tool-enzyme for genetic engineering is obtained.

Objective To observe the change of intima/media thickness ratio and expression of inflammatory factors in renal small artery of diabetic rats, and to explore the correlations of intim/media ratio with inflammatory factors and vascular lesions of diabetic nephropathy (DN) rats. Methods Seventy healthy SD rats were randomly divided into diabetic nephropathy group (DN, n=40) and normal control group (N, n=30). DN rat model was induced by intraperitoneal injection of streptozotocin (STZ). Thirty-five DN rats were successfully established. N group received same dose of citrate buffer. Rats were sacrificed after 4, 12, 24 weeks respectively.The intima/media thickness ratio in renal small artery was detected by immunofluorescence. The monocyte chemoattractant protein-1 (MCP-1) protein and mRNA expression of renal small artery were detected by immunohistochemistry and in-situ hybridization at each time point. Results Blood glucose and urine protein excretion (24 h) at different time points in DN group were significantly higher than those of N group (P＜0.05). From the 12th week, Scr, BUN, serum phosphorus were significantly higher than those of N group (P＜0.05). At the 4th week, renal small artery had the expression of MCP-1 protein and mRNA. The expression increased gradually with time, reached the highest at the 24th week, and was significantly higher than that of N group at each time point (P＜0.05). Immunofluorescence results showed that as compared to N group, in the first 4 weeks, intima/media thickness ratio in DN group was not different, at the 12th week the ratio was higher but without significant difference, at the 24th week the ratio was significantly higher (P＜0.05). Small artery intima/media thickness ratio of DN group was positively correlated with MCP-1, cholesterol, triglyceride, serum phosphate (r=0.742, P＜0.01; r=0.740, P＜0.01; r=0.829, P＜0.01; r=0.580, P＜0.01). Conclusions The arterioles intima/media thickness ratio of early DN is significantly correlated with MCP-1, lipids and phosphorus. MCP-1 may be involved in the DN vascular disease.