Objective: Machine learning (ML) has been reported to have better predictive capability than traditional statistical techniques. The aim of this study was to assess the efficacy of ML algorithms and logistic regression (LR) for predicting depressive symptoms during the COVID-19 pandemic. Methods: Analyses were carried out in a national cross-sectional study involving 21,916 participants. The ML algorithms in this study included random forest (RF), support vector machine (SVM), neural network (NN), and gradient boosting machine (GBM) methods. The performance indices were sensitivity, specificity, accuracy, precision, F1-score, and area under the receiver operating characteristic curve (AUC). Results: LR and NN had the best performance in terms of AUCs. The risk of overfitting was found to be negligible for most ML models except for RF, and GBM obtained the highest sensitivity, specificity, accuracy, precision, and F1-score. Therefore, LR, NN, and GBM models ranked among the best models. Conclusion Compared with ML models, LR model performed comparably to ML models in predicting depressive symptoms and identifying potential risk factors while also exhibiting a lower risk of overfitting.

Objective: Machine learning (ML) has been reported to have better predictive capability than traditional statistical techniques. The aim of this study was to assess the efficacy of ML algorithms and logistic regression (LR) for predicting depressive symptoms during the COVID-19 pandemic. Methods: Analyses were carried out in a national cross-sectional study involving 21,916 participants. The ML algorithms in this study included random forest (RF), support vector machine (SVM), neural network (NN), and gradient boosting machine (GBM) methods. The performance indices were sensitivity, specificity, accuracy, precision, F1-score, and area under the receiver operating characteristic curve (AUC). Results: LR and NN had the best performance in terms of AUCs. The risk of overfitting was found to be negligible for most ML models except for RF, and GBM obtained the highest sensitivity, specificity, accuracy, precision, and F1-score. Therefore, LR, NN, and GBM models ranked among the best models. Conclusion Compared with ML models, LR model performed comparably to ML models in predicting depressive symptoms and identifying potential risk factors while also exhibiting a lower risk of overfitting.

This study aims to explore the medication rules and mechanisms of treating chronic renal failure(CRF) by Jinling medical school based on data mining, network pharmacology, and experimental validation systematically and deeply. Firstly, the study selected the papers published by the inherited clinicians in Jinling medical school in Chinese journals using the subject headings named "traditional Chinese medicine(TCM) + chronic renal failure", "TCM + chronic renal inefficiency", or "TCM + consumptive disease" in China National Knowledge Infrastructure, Wanfang, and VIP Chinese Science and Technology Periodical Database and screened TCM formulas for treating CRF according to inclusion and exclusion criteria. The study analyzed the frequency of use of single TCM and the four properties, five tastes, channel tropism, and efficacy of TCM used with high frequency and performed association rule and clustering analysis, respectively. As a result, a total of 215 TCM formulas and 235 different single TCM were screened, respectively. The TCM used with high frequency included Astragali Radix, Rhei Radix et Rhizoma, Salviae Miltiorrhizae Radix et Rhizoma, Poria, and Atractylodis Macrocephalae Rhizoma(top 5). The single TCM characterized by "cold properties, sweet flavor, and restoring spleen channel" and the TCM with the efficacy of tonifying deficiency had the highest frequency of use, respectively. Then, the TCM with the rules of "blood-activating and stasis-removing" and "diuretic and dampness-penetrating" appeared. In addition, the core combination of TCM [(Hexin Formula, HXF)] included "Astragali Radix, Rhei Radix et Rhizoma, Poria, Salviae Miltiorrhizae Radix, and Angelicae Sinensis Radix". The network pharmacology analysis showed that HXF had 91 active compounds and 250 corresponding protein targets including prostaglandin-endoperoxide synthase 2(PTGS2), PTGS1, sodium voltage-gated channel alpha subunit 5(SCN5A), cholinergic receptor muscarinic 1(CHRM1), and heat shock protein 90 alpha family class A member 1(HSP90AA1)(top 5). Gene Ontology(GO) function analysis revealed that the core targets of HXF predominantly affected biological processes, cellular components, and molecular functions such as positive regulation of transcription by ribonucleic acid polymerase Ⅱ and DNA template transcription, formation of cytosol, nucleus, and plasma membrane, and identical protein binding and enzyme binding. Kyoto Encyclopedia of Genes and Genomes(KEGG) analysis revealed that CRF-related genes were involved in a variety of signaling pathways and cellular metabolic pathways, primarily involving "phosphatidylinositol 3-kinase(PI3K)-protein kinase B(Akt) pathway" and "advanced glycation end products-receptor for advanced glycation end products". Molecular docking results showed that the active components in HXF such as isomucronulatol 7-O-glucoside, betulinic acid, sitosterol, and przewaquinone B might be crucial in the treatment of CRF. Finally, a modified rat model with renal failure induced by adenine was used, and the in vivo experimental confirmation was performed based on the above-mentioned predictions. The results verify that HXF can regulate mitochondrial autophagy in the kidneys and the PI3K-Akt-mammalian target of rapamycin(mTOR) signaling pathway activation at upstream, so as to alleviate renal tubulointerstitial fibrosis and then delay the progression of CRF.
Data Mining ; Drugs, Chinese Herbal/chemistry* ; Network Pharmacology ; Humans ; Kidney Failure, Chronic/metabolism* ; Medicine, Chinese Traditional ; China

This study aims to explore the mechanism through which Yishen Jiangtang Decoction(YSJTD) regulates endoplasmic reticulum stress(ERS)-mediated NOD-like receptor thermal protein domain associated protein 3(NLRP3) inflammasome to improve diabetic nephropathy(DN) in db/db mice. Thirty db/db mice were randomly divided into the model group, YSJTD group, ERS inhibitor 4-phenylbutyric acid(4-PBA) group, with 10 mice in each group. Additionally, 10 db/m mice were selected as the control group. The YSJTD group was orally administered YSJTD at a dose of 0.01 mL·g~(-1), the 4-PBA group was orally administered 4-PBA at a dose of 0.5 mg·g~(-1), and the control and model groups were given an equal volume of carboxylmethyl cellulose sodium. The treatments were administered once daily for 8 weeks. Food intake, water consumption, and body weight were recorded every 2 weeks. After the intervention, fasting blood glucose(FBG), glycosylated hemoglobin(HbA1c), urine microalbumin(U-mALB), 24-hour urine volume, serum creatinine(Scr), and blood urea nitrogen(BUN) were measured. Inflammatory markers interleukin-1β(IL-1β) and interleukin-18(IL-18) were detected using the enzyme-linked immunosorbent assay(ELISA). Renal pathology was assessed through hematoxylin-eosin(HE), periodic acid-Schiff(PAS), and Masson staining, and transmission electron microscopy(TEM). Western blot was used to detect the expression levels of glucose-regulated protein 78(GRP78), C/EBP homologous protein(CHOP), NLRP3, apoptosis-associated speck-like protein containing CARD(ASC), cysteinyl aspartate-specific proteinase(caspase-1), and gasdermin D(GSDMD) in kidney tissues. The results showed that compared to the control group, the model group exhibited poor general condition, increased weight and food and water intake, and significantly higher levels of FBG, HbA1c, U-mALB, kidney index, 24-hour urine volume, IL-1β, and IL-18. Compared to the model group, the YSJTD and 4-PBA groups showed improved general condition, increased body weight, decreased food intake, and lower levels of FBG, U-mALB, kidney index, 24-hour urine volume, and IL-1β. Specifically, the YSJTD group showed a significant reduction in IL-18 levels compared to the model group, while the 4-PBA group exhibited decreased water intake and HbA1c levels compared to the model group. Although there was a decreasing trend in water intake and HbA1c in the YSJTD group, the differences were not statistically significant. No significant differences were observed in BUN, Scr, and kidney weight among the groups. Renal pathology revealed that the model group exhibited more severe renal damage compared to the control group. Kidney sections from the model group showed diffuse mesangial proliferation in the glomeruli, tubular edema, tubular dilation, significant inflammatory cell infiltration in the interstitium, and increased glycogen staining and blue collagen deposition in the basement membrane. In contrast, the YSJTD and 4-PBA groups showed varying degrees of improvement in renal damage, glycogen staining, and collagen deposition, with the YSJTD group showing more significant improvements. TEM analysis indicated that the model group had extensive cytoplasmic edema, homogeneous thickening of the basement membrane, fewer foot processes, and widening of fused foot processes. In the YSJTD and 4-PBA groups, cytoplasmic swelling of renal tissues was reduced, the basement membrane remained intact and uniform, and foot process fusion improved.Western blot results indicated that compared to the control group, the model group showed upregulation of GRP78, CHOP, GSDMD, NLRP3, ASC, and caspase-1 expression. In contrast, both the YSJTD and 4-PBA groups showed downregulation of these markers compared to the model group. These findings suggest that YSJTD exerts a protective effect against DN by alleviating NLRP3 inflammasome activation through the inhibition of ERS, thereby improving the inflammatory response in db/db DN mice.
Animals ; Endoplasmic Reticulum Stress/drug effects* ; Diabetic Nephropathies/metabolism* ; NLR Family, Pyrin Domain-Containing 3 Protein/genetics* ; Drugs, Chinese Herbal/administration & dosage* ; Mice ; Inflammasomes/drug effects* ; Male ; Kidney/pathology* ; Endoplasmic Reticulum Chaperone BiP ; Humans ; Interleukin-18/genetics* ; Mice, Inbred C57BL

Extensive epigenetic reprogramming involves in memory CD8+ T-cell differentiation. The elaborate epigenetic rewiring underlying the heterogeneous functional states of CD8+ T cells remains hidden. Here, we profile single-cell chromatin accessibility and map enhancer-promoter interactomes to characterize the differentiation trajectory of memory CD8+ T cells. We reveal that under distinct epigenetic regulations, the early activated CD8+ T cells divergently originated for short-lived effector and memory precursor effector cells. We also uncover a defined epigenetic rewiring leading to the conversion from effector memory to central memory cells during memory formation. Additionally, we illustrate chromatin regulatory mechanisms underlying long-lasting versus transient transcription regulation during memory differentiation. Finally, we confirm the essential roles of Sox4 and Nrf2 in developing memory precursor effector and effector memory cells, respectively, and validate cell state-specific enhancers in regulating Il7r using CRISPR-Cas9. Our data pave the way for understanding the mechanism underlying epigenetic memory formation in CD8+ T-cell differentiation.
CD8-Positive T-Lymphocytes/metabolism* ; Cell Differentiation ; Chromatin/immunology* ; Animals ; Mice ; Immunologic Memory ; Epigenesis, Genetic ; SOXC Transcription Factors/immunology* ; NF-E2-Related Factor 2/immunology* ; Mice, Inbred C57BL ; Gene Regulatory Networks ; Enhancer Elements, Genetic

Hepatocellular carcinoma (HCC) expresses abundant glycolytic enzymes and displays comprehensive glucose metabolism reprogramming. Aldolase A (ALDOA) plays a prominent role in glycolysis; however, little is known about its role in HCC development. In the present study, we aim to explore how ALDOA is involved in HCC proliferation. HCC proliferation was markedly suppressed both in vitro and in vivo following ALDOA knockout, which is consistent with ALDOA overexpression encouraging HCC proliferation. Mechanistically, ALDOA knockout partially limits the glycolytic flux in HCC cells. Meanwhile, ALDOA translocated to nuclei and directly interacted with c-Jun to facilitate its Thr93 phosphorylation by P21-activated protein kinase; ALDOA knockout markedly diminished c-Jun Thr93 phosphorylation and then dampened c-Jun transcription function. A crucial site Y364 mutation in ALDOA disrupted its interaction with c-Jun, and Y364S ALDOA expression failed to rescue cell proliferation in ALDOA deletion cells. In HCC patients, the expression level of ALDOA was correlated with the phosphorylation level of c-Jun (Thr93) and poor prognosis. Remarkably, hepatic ALDOA was significantly upregulated in the promotion and progression stages of diethylnitrosamine-induced HCC models, and the knockdown of A ldoa strikingly decreased HCC development in vivo. Our study demonstrated that ALDOA is a vital driver for HCC development by activating c-Jun-mediated oncogene transcription, opening additional avenues for anti-cancer therapies.

Objective: Machine learning (ML) has been reported to have better predictive capability than traditional statistical techniques. The aim of this study was to assess the efficacy of ML algorithms and logistic regression (LR) for predicting depressive symptoms during the COVID-19 pandemic. Methods: Analyses were carried out in a national cross-sectional study involving 21,916 participants. The ML algorithms in this study included random forest (RF), support vector machine (SVM), neural network (NN), and gradient boosting machine (GBM) methods. The performance indices were sensitivity, specificity, accuracy, precision, F1-score, and area under the receiver operating characteristic curve (AUC). Results: LR and NN had the best performance in terms of AUCs. The risk of overfitting was found to be negligible for most ML models except for RF, and GBM obtained the highest sensitivity, specificity, accuracy, precision, and F1-score. Therefore, LR, NN, and GBM models ranked among the best models. Conclusion Compared with ML models, LR model performed comparably to ML models in predicting depressive symptoms and identifying potential risk factors while also exhibiting a lower risk of overfitting.

Objective: Machine learning (ML) has been reported to have better predictive capability than traditional statistical techniques. The aim of this study was to assess the efficacy of ML algorithms and logistic regression (LR) for predicting depressive symptoms during the COVID-19 pandemic. Methods: Analyses were carried out in a national cross-sectional study involving 21,916 participants. The ML algorithms in this study included random forest (RF), support vector machine (SVM), neural network (NN), and gradient boosting machine (GBM) methods. The performance indices were sensitivity, specificity, accuracy, precision, F1-score, and area under the receiver operating characteristic curve (AUC). Results: LR and NN had the best performance in terms of AUCs. The risk of overfitting was found to be negligible for most ML models except for RF, and GBM obtained the highest sensitivity, specificity, accuracy, precision, and F1-score. Therefore, LR, NN, and GBM models ranked among the best models. Conclusion Compared with ML models, LR model performed comparably to ML models in predicting depressive symptoms and identifying potential risk factors while also exhibiting a lower risk of overfitting.

Objective: Machine learning (ML) has been reported to have better predictive capability than traditional statistical techniques. The aim of this study was to assess the efficacy of ML algorithms and logistic regression (LR) for predicting depressive symptoms during the COVID-19 pandemic. Methods: Analyses were carried out in a national cross-sectional study involving 21,916 participants. The ML algorithms in this study included random forest (RF), support vector machine (SVM), neural network (NN), and gradient boosting machine (GBM) methods. The performance indices were sensitivity, specificity, accuracy, precision, F1-score, and area under the receiver operating characteristic curve (AUC). Results: LR and NN had the best performance in terms of AUCs. The risk of overfitting was found to be negligible for most ML models except for RF, and GBM obtained the highest sensitivity, specificity, accuracy, precision, and F1-score. Therefore, LR, NN, and GBM models ranked among the best models. Conclusion Compared with ML models, LR model performed comparably to ML models in predicting depressive symptoms and identifying potential risk factors while also exhibiting a lower risk of overfitting.

OBJECTIVE To construct and apply drug utilization evaluation （DUE） criteria for low-dose rivaroxaban in atherosclerotic cardiovascular disease （ASCVD） based on the dual pathway inhibition （DPI） antithrombotic therapy scheme， to promote clinical rational drug use. METHODS Based on the instructions and relevant guidelines of low-dose rivaroxaban （2.5 mg， bid）， the Delphi method was used to establish the DUE criteria for low-dose rivaroxaban used in ASCVD. Weighted technique for order preference by similarity to an ideal solution method was used to determine the relative weights of each evaluation index， and the rationality of the filing medical records of discharged patients using low-dose rivaroxaban for ASCVD at Anqing Municipal Hospital from February 2024 to January 2025 was evaluated. RESULTS The established DUE criteria included 3 primary indicators （medication indications， medication process， medication results） and 11 secondary indicators （such as indications， contraindications， etc.）. The higher weighted secondary indicators being contraindications （0.117 9） and indications （0.112 1）. A total of 265 medical records were included for evaluation. The evaluation results showed that 192 cases （72.45%） had reasonable medical records， 69 cases （26.04%） had basic reasonable medical records， and 4 cases （1.51%） had unreasonable medical records； unreasonable types mainly included inappropriate combination therapy， inappropriate usage and dosage， inappropriate post- medication monitoring， and inappropriate drug switching， etc. CONCLUSIONS This study establishes a DUE criteria for low-dose rivaroxaban in ASCVD based on the DPI antithrombotic treatment regimen， and the evaluation results are intuitive， reliable， and quantifiable. The use of low-dose rivaroxaban in ASCVD patients in our hospital is relatively reasonable， but further management needs to be strengthened.