Network pharmacology and Mendelian randomization analysis to unravel the mechanism of Qingwen Baidu Decoction in sepsis treatment

The aim of this study was to evaluate the potential components and targets of Qingwen Baidu Decoction (QBD) for sepsis by integrating a comprehensive strategy of network pharmacology and Mendelian randomization (MR). Systematically search the Traditional Chinese Medicine Systems Pharmacology, HERB, and Stitch databases for potential active ingredients and therapeutic targets of QBD. GeneCards, and DisGeNET databases were used to find the target of sepsis. We analyze the active ingredients and disease targets of drugs through network pharmacology, combined with pathway enrichment, protein–protein interaction and “Drug-component-target” network analysis to explore molecular mechanisms. MR was performed to assess the causal relationship between key targets and sepsis risk. The eQTLGen Consortium (sample sizes: 31,684) provided data on cis-expression quantitative trait loci (exposure). The summary data on sepsis (outcome) from genome-wide association studies were gathered from the IEU OpenGWAS project (sample sizes: 486,484). We built a target interaction network between the probable active ingredient targets of QBD and the therapeutic targets of sepsis. We identified 216 active ingredients in QBD, corresponding to 626 drug targets and 2575 sepsis-related targets, with 282 overlapping targets. Key active ingredients include quercetin, kaempferol, luteolin, wogonin, and naringenin. Protein–protein interaction network analysis revealed 17 hub targets. Gene ontology analysis demonstrated significant involvement in biological processes such as inflammatory response, while Kyoto encyclopedia of genes and genomes pathway analysis highlighted pathways including PI3K-Akt, TNF, HIF-1, and interleukin-17 signaling. MR analysis suggested a potential genetic association between STAT1 and increased sepsis risk (OR = 1.51, P = .024). STAT1 was predicted as a potential candidate target linking QBD to sepsis based on integrated network pharmacology and MR, but this finding requires experimental validation and does not establish therapeutic causality.