Objective To investigate the molecular mechanism of hepatic fibrosis (HF) regulation by liver sinusoidal endothelial cells (LSECs) via endothelial mesenchymal transition (EnMT), and to predict the natural active components using bioinformatics, machine learning, and cellular experiments. Methods HF and EnMT gene matrices were obtained and the intersecting genes were extracted and enriched using Limma difference analysis and weighted gene co-expression network analysis (WGCNA). The diagnostic genes were screened using a combination of random forest method, support vector machine-recursive feature elimination and network topology analysis, and immune infiltration analysis and prediction of natural active ingredients were performed. The expression of diagnostic genes and the pharmacological effects of the predicted ingredients were finally verified by cellular experiments. Results Differentialanalysis yielded 3034 EnMT-associated and 4133 HF-associated differential genes. WGCNA analysis yielded 4589 EnMT-associated Hub genes and 763 HF-associated Hub genes. Thirty-eight intersecting genes were extracted, which were mainly enriched in the pathways of basement membrane and extracellular matrix receptor interaction. Four diagnostic genes, CFP, COL4A2, ITGA1, and GRPEL1, were screened by multidimensional analysis. Immune infiltration analysis showed that the diagnostic genes were closely associated with mast cell resting state, memory B cells, and memory CD4⁺ T cells. Reverse transcription-polymerase chain reaction analysis showed significantly increased mRNA expression levels of the four diagnostic genes in the Jagged1-induced model group (P<0.05). The predicted components, sterol, kaempferol, and quercetin, all had good binding activities with the diagnostic genes. Enzyme-linked immunosorbent assay result confirmed that all three active components significantly reduced the expression of collagen type IV α2 chain protein in Jagged1-induced LSECs, with quercetin having the most significant effect (P<0.01). Conclusions This study elucidated the molecular mechanism of hepatic sinusoidal endothelial cells involved in the pathological process of HF through mesenchymal transition. We also propose a diagnostic marker system including CFP, COL4A2, ITGA1, and GRPEL1 as core genes. The result also suggest that natural active ingredients, such as quercetin, may exert anti-HF pharmacological effects by targeting these diagnostic genes.