1.Construction and validation of predictive models for intravenous immunoglobulin–resistant Kawasaki disease using an interpretable machine learning approach
Linfan DENG ; Jian ZHAO ; Ting WANG ; Bin LIU ; Jun JIANG ; Peng JIA ; Dong LIU ; Gang LI
Clinical and Experimental Pediatrics 2024;67(8):405-414
Background:
Intravenous immunoglobulin (IVIG)-resistant Kawasaki disease is associated with coronary artery lesion development.Purpose: This study aimed to explore the factors associated with IVIG-resistance and construct and validate an interpretable machine learning (ML) prediction model in clinical practice.
Methods:
Between December 2014 and November 2022, 602 patients were screened and risk factors for IVIG-resistance investigated. Five ML models are used to establish an optimal prediction model. The SHapley Additive exPlanations (SHAP) method was used to interpret the ML model.
Results:
Na+, hemoglobin (Hb), C-reactive protein (CRP), and globulin were independent risk factors for IVIG-resistance. A nonlinear relationship was identified between globulin level and IVIG-resistance. The XGBoost model exhibited excellent performance, with an area under the receiver operating characteristic curve of 0.821, accuracy of 0.748, sensitivity of 0.889, and specificity of 0.683 in the testing set. The XGBoost model was interpreted globally and locally using the SHAP method.
Conclusion
Na+, Hb, CRP, and globulin levels were independently associated with IVIG-resistance. Our findings demonstrate that ML models can reliably predict IVIG-resistance. Moreover, use of the SHAP method to interpret the established XGBoost model's findings would provide evidence of IVIG-resistance and guide the individualized treatment of Kawasaki disease.
2.Construction and validation of predictive models for intravenous immunoglobulin–resistant Kawasaki disease using an interpretable machine learning approach
Linfan DENG ; Jian ZHAO ; Ting WANG ; Bin LIU ; Jun JIANG ; Peng JIA ; Dong LIU ; Gang LI
Clinical and Experimental Pediatrics 2024;67(8):405-414
Background:
Intravenous immunoglobulin (IVIG)-resistant Kawasaki disease is associated with coronary artery lesion development.Purpose: This study aimed to explore the factors associated with IVIG-resistance and construct and validate an interpretable machine learning (ML) prediction model in clinical practice.
Methods:
Between December 2014 and November 2022, 602 patients were screened and risk factors for IVIG-resistance investigated. Five ML models are used to establish an optimal prediction model. The SHapley Additive exPlanations (SHAP) method was used to interpret the ML model.
Results:
Na+, hemoglobin (Hb), C-reactive protein (CRP), and globulin were independent risk factors for IVIG-resistance. A nonlinear relationship was identified between globulin level and IVIG-resistance. The XGBoost model exhibited excellent performance, with an area under the receiver operating characteristic curve of 0.821, accuracy of 0.748, sensitivity of 0.889, and specificity of 0.683 in the testing set. The XGBoost model was interpreted globally and locally using the SHAP method.
Conclusion
Na+, Hb, CRP, and globulin levels were independently associated with IVIG-resistance. Our findings demonstrate that ML models can reliably predict IVIG-resistance. Moreover, use of the SHAP method to interpret the established XGBoost model's findings would provide evidence of IVIG-resistance and guide the individualized treatment of Kawasaki disease.
3.Construction and validation of predictive models for intravenous immunoglobulin–resistant Kawasaki disease using an interpretable machine learning approach
Linfan DENG ; Jian ZHAO ; Ting WANG ; Bin LIU ; Jun JIANG ; Peng JIA ; Dong LIU ; Gang LI
Clinical and Experimental Pediatrics 2024;67(8):405-414
Background:
Intravenous immunoglobulin (IVIG)-resistant Kawasaki disease is associated with coronary artery lesion development.Purpose: This study aimed to explore the factors associated with IVIG-resistance and construct and validate an interpretable machine learning (ML) prediction model in clinical practice.
Methods:
Between December 2014 and November 2022, 602 patients were screened and risk factors for IVIG-resistance investigated. Five ML models are used to establish an optimal prediction model. The SHapley Additive exPlanations (SHAP) method was used to interpret the ML model.
Results:
Na+, hemoglobin (Hb), C-reactive protein (CRP), and globulin were independent risk factors for IVIG-resistance. A nonlinear relationship was identified between globulin level and IVIG-resistance. The XGBoost model exhibited excellent performance, with an area under the receiver operating characteristic curve of 0.821, accuracy of 0.748, sensitivity of 0.889, and specificity of 0.683 in the testing set. The XGBoost model was interpreted globally and locally using the SHAP method.
Conclusion
Na+, Hb, CRP, and globulin levels were independently associated with IVIG-resistance. Our findings demonstrate that ML models can reliably predict IVIG-resistance. Moreover, use of the SHAP method to interpret the established XGBoost model's findings would provide evidence of IVIG-resistance and guide the individualized treatment of Kawasaki disease.
4.Construction and validation of predictive models for intravenous immunoglobulin–resistant Kawasaki disease using an interpretable machine learning approach
Linfan DENG ; Jian ZHAO ; Ting WANG ; Bin LIU ; Jun JIANG ; Peng JIA ; Dong LIU ; Gang LI
Clinical and Experimental Pediatrics 2024;67(8):405-414
Background:
Intravenous immunoglobulin (IVIG)-resistant Kawasaki disease is associated with coronary artery lesion development.Purpose: This study aimed to explore the factors associated with IVIG-resistance and construct and validate an interpretable machine learning (ML) prediction model in clinical practice.
Methods:
Between December 2014 and November 2022, 602 patients were screened and risk factors for IVIG-resistance investigated. Five ML models are used to establish an optimal prediction model. The SHapley Additive exPlanations (SHAP) method was used to interpret the ML model.
Results:
Na+, hemoglobin (Hb), C-reactive protein (CRP), and globulin were independent risk factors for IVIG-resistance. A nonlinear relationship was identified between globulin level and IVIG-resistance. The XGBoost model exhibited excellent performance, with an area under the receiver operating characteristic curve of 0.821, accuracy of 0.748, sensitivity of 0.889, and specificity of 0.683 in the testing set. The XGBoost model was interpreted globally and locally using the SHAP method.
Conclusion
Na+, Hb, CRP, and globulin levels were independently associated with IVIG-resistance. Our findings demonstrate that ML models can reliably predict IVIG-resistance. Moreover, use of the SHAP method to interpret the established XGBoost model's findings would provide evidence of IVIG-resistance and guide the individualized treatment of Kawasaki disease.
5.Research progress in clinical application and pharmacological effect of Yiyi Fuzi Baijiang Powder.
Liu-Yi WEI ; Er-Wei HAO ; Xiao-Tao HOU ; Liu-Zhi WEI ; Zhe QIN ; Xiao YANG ; Qian XIAO ; Shi-du YAN ; Jia-Gang DENG
China Journal of Chinese Materia Medica 2023;48(18):4893-4901
Yiyi Fuzi Baijiang Powder(YFBP), originating from Synopsis of the Golden Chamber, is a classic prescription composed of Coicis Semen, Aconiti Lateralis Radix Praeparata, and Patriniae Herba for the treatment of abscesses and pus discharge. This article presented a systematic analysis of the clinical application of YFBP, including the indicated diseases, the number of cases, efficacy, dosage, administration methods, and compatibility with other drugs. The analysis reveals that YFBP has a wide range of clinical applications. It is commonly used, often with modifications or in combination with western medicine, for diseases in the fields of gastroente-rology, gynecology, urology, dermatology, and others. And most of the Traditional Chinese Medicine(TCM) evidence involved in these diseases are damp-heat evudence. The prescription shows rich variations in clinical administration methods, and most of which are the treatment of aqueous decoction of it. The therapeutic effect is also significant, and the total effective rate of clinical treatment is re-latively high. Additionally, this article summarized the pharmacological research on YFBP and found that it possessed various pharmacological effects, including anti-inflammatory, antioxidant, anticancer, and immune-modulating properties. Finally, correlation analysis was conducted on the main diseases, TCM types, prescription doses, pharmacological effects and action targets of YFBP, which to show the relationship between these five aspects in a visual form, reflecting the relationship between its clinical application and modern pharmacological effects. These findings provide a reference basis for further development and research on YFBP.
Powders
;
Drugs, Chinese Herbal/pharmacology*
;
Medicine, Chinese Traditional
;
Diterpenes
;
Aconitum
6.Mechanism of Huangjing Qianshi Decoction in treatment of prediabetic mice based on transcriptome sequencing.
Jia-Luo CAI ; Yi-Lin ZHU ; Xiao-Ping LI ; Xin-Hua XIA ; Gui-Ming DENG ; Qiao-Zhen TONG ; Gang-Qiang YI ; Bo CHENG
China Journal of Chinese Materia Medica 2023;48(4):1032-1042
Based on transcriptome sequencing technology, the mouse model of prediabetes treated with Huangjing Qianshi Decoction was sequenced to explore the possible mechanism of treating prediabetes. First of all, transcriptome sequencing was performed on the normal BKS-DB mouse group, the prediabetic model group, and the Huangjing Qianshi Decoction treatment group(treatment group) to obtain differentially expressed genes in the skeletal muscle samples of mice. The serum biochemical indexes were detected in each group to screen out the core genes of Huangjing Qianshi Decoction in prediabetes. Gene Ontology(GO) database and Kyoto Encyclopedia of Genes and Genomes(KEGG) database were used to conduct signaling pathway enrichment analysis of differentially expressed genes, and real-time quantitative polymerase chain reaction(RT-qPCR) was used to verify them. The results showed that the levels of fasting blood glucose(FBG), fasting insulin(FINS), insulin resistance index(HOMA-IR), total cholesterol(TC), triglycerides(TG), and low-density lipoprotein cholesterol(LDL-C) in the mouse model were significantly decreased after treatment with Huangjing Qianshi Decoction. In the results of differential gene screening, there were 1 666 differentially expressed genes in the model group as compared with the normal group, and there were 971 differentially expressed genes in the treatment group as compared with the model group. Among them, interleukin-6(IL-6) and NR3C2 genes, which were closely related to the regulation of insulin resis-tance function, were significantly up-regulated between the model group and the normal group, and vascular endothelial growth factor A(VEGFA) genes were significantly down-regulated between the model group and the normal group. However, the expression results of IL-6, NR3C2, and VEGFA genes were adverse between the treatment group and the model group. GO functional enrichment analysis found that the biological process annotation mainly focused on cell synthesis, cycle, and metabolism; cell component annotation mainly focused on organelles and internal components; and molecular function annotation mainly focused on binding molecular functions. KEGG pathway enrichment analysis found that it involved the protein tyrosine kinase 6(PTK6) pathway, CD28-dependent phosphoinositide 3-kinase/protein kinase B(PI3K/AKT) pathway, p53 pathway, etc. Therefore, Huangjing Qianshi Decoction can improve the state of prediabetes, and the mechanism may be related to cell cycle and apoptosis, PI3K/AKT pathway, p53 pathway, and other biological pathways regulated by IL-6, NR3C2, and VEGFA.
Animals
;
Mice
;
Proto-Oncogene Proteins c-akt
;
Phosphatidylinositol 3-Kinases
;
Prediabetic State
;
Vascular Endothelial Growth Factor A
;
Interleukin-6
;
Transcriptome
;
Tumor Suppressor Protein p53
;
Insulin
;
Cholesterol
7.Medication Rules of Clinical Prescriptions Containing Citri Sarcodactylis Fructus Based on Data Mining
Qian XIAO ; Er-wei HAO ; Zheng-cai DU ; Dong-yang WU ; Ling-ling LIANG ; Ze-yu LI ; Xiao-tao HOU ; Jia-gang DENG
Chinese Journal of Experimental Traditional Medical Formulae 2022;28(8):194-203
ObjectiveTo reveal the medication rules of the clinical prescriptions containing Citri Sarcodactylis Fructus, and to provide a basis for the modern clinical application of Citri Sarcodactylis Fructus, the development of health products and the research and development of new drugs. MethodThe clinical prescriptions containing Citri Sarcodactylis Fructus were retrieved from CNKI, Wanfang Data, and VIP, and then a basic database of prescriptions was established via the traditional Chinese medicine inheritance auxiliary platform (V2.5), IBM SPSS V20, and Excel (Microsoft Office 2016). The frequency and association rules of the medicines in prescriptions (compatible medicines) and the corresponding syndromes were then mined. ResultThe prescriptions were screened according to the inclusion and exclusion criteria. A total of 458 clinical prescriptions containing Citri Sarcodactylis Fructus were collected, involving 388 Chinese medicines, and the total frequency of medicines reached 6 199. The core compatible medicines (frequency > 130) of Citri Sarcodactylis Fructus included Poria (frequency of 222), Atractylodis Macrocephalae Rhizoma (217), Paeoniae Radix Alba (196), Bupleuri Radix (159), and Citri Reticulatae Pericarpium (142). The Citri Sarcodactylis Fructus-compatible medicines with frequency > 49 were selected for further analysis, which included 34 medicines with the cumulative frequency of 3 131 (50.51% of the total frequency). These medicines mainly have the functions of tonifying Qi, invigorating Qi, tonifying blood, alleviating edema and promoting urination, promoting digestion, and activating blood and relieving pain. They are mainly warm, cold, or mild-natured, taste bitter, sweet, or acrid, and have the tropism in the spleen, liver, stomach, or lung meridians. The association rule analysis demonstrated that 14 medicine combinations were commonly used, and the core combinations were Poria-Citri Sarcodactylis Fructus, Atractylodis Macrocephalae Rhizoma-Citri Sarcodactylis Fructus, Paeoniae Radix Alba-Citri Sarcodactylis Fructus, Bupleuri Radix-Citri Sarcodactylis Fructus, Atractylodis Macrocephalae Rhizoma-Poria-Citri Sarcodactylis Fructus, and Citri Reticulatae Pericarpium-Citri Sarcodactylis Fructus. The clinical prescriptions containing Citri Sarcodactylis Fructus were mainly used to treat 52 diseases corresponding to 11 types of traditional Chinese medicine (TCM) syndromes. Three representative syndrome types, including spleen and stomach syndromes, Qi-blood-body fluid syndromes, and gynecological syndromes were selected for further association rule analysis. In the treatment of spleen and stomach syndromes, the core compatible drugs were Atractylodis Macrocephalae Rhizoma, Poria, Paeoniae Radix Alba, Bupleuri Radix, Citri Reticulatae Pericarpium, and Pinelliae Ehizoma, which, together with Citri Sarcodactylis Fructus, formed 25 commonly used medicine combinations (16 combinations composed of 2 medicines and 9 combinations composed of 3 medicines). In the treatment of Qi-blood-body fluid syndromes, the core compatible drugs were Poria, Paeoniae Radix Alba, Atractylodis Macrocephalae Rhizoma, Astragali Radix, Hordei Fructus Germinatus, and Bupleuri Radix, which, together with Citri Sarcodactylis Fructus, formed 23 common medicine combinations (17 combinations composed of 2 medicines, 5 combinations composed of 3 medicines, and 1 combination composed of 4 medicines). In the treatment of gynecological syndromes, the core compatible medicines were Atractylodis Macrocephalae Rhizoma, Paeoniae Radix Alba, Angelicae Sinensis Radix, Astragali Radix, Cyperi Rhizoma, and Poria, which constituted 25 common medicine combinations (15 combinations composed of 2 medicines and 10 combinations composed of 3 medicines). ConclusionWe employed the traditional Chinese medicine(TCM) inheritance auxiliary platform to explore the compatibility of Citri Sarcodactylis Fructus-containing clinical prescriptions and the corresponding TCM syndromes, which intuitively showcased the medication rules of Citri Sarcodactylis Fructus. Specifically, Citri Sarcodactylis Fructus was mainly combined with the medicines for tonifying Qi, invigorating Qi, tonifying blood, alleviating edema and promoting urination, promoting digestion, and activating blood and relieving pain to treat different TCM syndromes. While soothing liver, regulating Qi, harmonizing stomach, and relieving pain, the combinations tonify and activate blood, invigorate spleen, and resolve dampness. The findings are of great significance to the rational application of Citri Sarcodactylis Fructus, the development of health food, and the research of new drugs and will bolster the development of Chinese medicine industry.
8.Action Mechanism of Chinese Medicine Polysaccharides in Prevention and Treatment Diabetes and Its Complications:A Review
Feng CHEN ; Fan ZHANG ; Er-wei HAO ; Liu-ying MO ; Jin-ling XIE ; Ze-yu LI ; Xing-yuan WANG ; Rui CAO ; Xiao-tao HOU ; Jia-gang DENG
Chinese Journal of Experimental Traditional Medical Formulae 2022;28(12):256-266
Diabetes is a metabolic disease mainly characterized by hyperglycemia due to inadequate insulin secretion. And persistent hyperglycemia can cause chronic damage or dysfunction of eyes, kidneys, heart, blood vessels and nerves. Polysaccharides are high molecular carbohydrates polymerized by glycosidic bonds from more than 10 monosaccharide molecules of the same or different types. They have the advantages of wide sources, high safety and low toxic and so on. As one of the important effective components of traditional Chinese medicine, polysaccharides have biological activities such as immune regulation, anti-oxidation, anti-tumor, lowering blood sugar and so on. The structure is directly related to biological activities, and the advanced structure of polysaccharides is based on the primary structure. Exploring the primary structure of polysaccharides is the key task of lowering blood sugar and improving diabetic complications. This paper summarizes the monosaccharide composition of the primary structure of Chinese medicine polysaccharides, and the mechanism of Chinese medicine polysaccharides improving diabetes is emphasized by increasing the secretion and release of insulin, increasing the islet β cell number, upregulating insulin receptor level, improving glucose and lipid metabolism, inhibiting inflammatory response, improving oxidative stress and regulating phosphatidylinositol-3-kinase(PI3K)/protein kinase B (Akt), mitogen activated protein kinase, cyclic adenosine monophosphate(cAMP)/protein kinase A(PKA) and adenosine monophosphate activated protein kinase(AMPK) signaling pathways. At the same time, we also summarized the prevention and treatment of Chinese medicine polysaccharides in diabetic nephropathy, diabetic cardiomyopathy, diabetic ophthalmopathy and diabetic peripheral neuropathy, in order to provide a theoretical basis for new drug development and clinical application of Chinese medicine polysaccharides in the intervention of diabetes and its complications.
9.Mechanism of Huangjing Qianshi Decoction in treatment of prediabetes based on network pharmacology and molecular docking.
Jia-Luo CAI ; Xiao-Ping LI ; Yi-Lin ZHU ; Gui-Ming DENG ; Lei YANG ; Xin-Hua XIA ; Gang-Qiang YI ; Xin-Yu CHEN
China Journal of Chinese Materia Medica 2022;47(4):1039-1050
This study analyzed the molecular mechanism of Huangjing Qianshi Decoction(HQD) in the treatment of prediabetes based on network pharmacology and molecular docking. The active components of HQD were identified and screened based on Traditional Chinese Medicine Systems Pharmacology Database and Analysis Platform(TCMSP, http://Lsp.nwu.edu.cn/tcmsp.php) and then the targets of the components and the genes related to prediabetes were retrieved, followed by identifying the common targets of the decoction and the disease. The medicinal component-target network was constructed by Cytoscape to screen key components. The protein-protein interaction(PPI) network was established by STRING and hub genes were identified by Cytoscape-CytoNCA, followed by Gene Ontology(GO) term enrichment and Kyoto Encyclopedia of Genes and Genomes(KEGG) of the hub genes with R-clusterProfi-ler. Thereby, the possible signaling pathways were predicted and the molecular mechanism was deduced. A total of 79 active components of HQD and 785 diabetes-related targets of the components were screened out. The hub genes mainly involved the GO terms of tricarboxylic acid cycle, peptide binding, amide binding, hydrolase activity, and kinase activity regulation, and the KEGG pathways of AGE-RAGE signaling pathway, TNF signaling pathway, AMPK signaling pathway, IL-17 signaling pathway, and insulin signaling pathway. Western blot result showed that HQD-containing serum significantly reduced the expression of AKT1, AGE, and RAGE proteins in insulin resistance model cells. HQD's treatment of prediabetes is characterized by multiple pathways, multiple targets, and multiple levels. The main mechanism is that the components zhonghualiaoine, baicalein, kaempferol, and luteolin act on AKT1 and inhibit the AGE-RAGE axis.
Drugs, Chinese Herbal/pharmacology*
;
Humans
;
Medicine, Chinese Traditional
;
Molecular Docking Simulation
;
Network Pharmacology
;
Prediabetic State/genetics*
10.Research progress of Curcuma kwangsiensis root tubers and analysis of liver protection and anti-tumor mechanisms based on Q-markers.
Ze-Yu LI ; Er-Wei HAO ; Zheng-Cai DU ; Rui CAO ; Feng CHEN ; Liu-Ying MO ; Dong-Yang WU ; Xiao-Tao HOU ; Jia-Gang DENG
China Journal of Chinese Materia Medica 2022;47(7):1739-1753
Curcuma kwangsiensis root tuber is a widely used genuine medicinal material in Guangxi, with the main active components of terpenoids and curcumins. It has the effects of promoting blood circulation to relieve pain, moving Qi to relieve depression, clearing heart and cooling blood, promoting gallbladder function and anti-icterus. Modern research has proved its functions in liver protection, anti-tumor, anti-oxidation, blood lipid reduction and immunosuppression. Considering the research progress of C. kwangsiensis root tubers and the core concept of quality marker(Q-marker), we predicted the Q-markers of C. kwangsiensis root tubers from plant phylogeny, chemical component specificity, traditional pharmacodynamic properties, new pharmacodynamic uses, chemical component measurability, processing methods, compatibility, and components migrating to blood. Curcumin, curcumol, curcumadiol, curcumenol, curdione, germacrone, and β-elemene may be the possible Q-markers. Based on the predicted Q-markers, the mechanisms of the liver-protecting and anti-tumor activities of C. kwangsiensis root tubers were analyzed. AKT1, IL6, EGFR, and STAT3 were identified as the key targets, and neuroactive ligand-receptor interaction signaling pathway, nitrogen metabolism pathway, cancer pathway, and hepatitis B pathway were the major involved pathways. This review provides a basis for the quality evaluation and product development of C. kwangsiensis root tubers and gives insights into the research on Chinese medicinal materials.
China
;
Curcuma/chemistry*
;
Humans
;
Liver
;
Neoplasms
;
Terpenes/pharmacology*

Result Analysis
Print
Save
E-mail