Ovarian cancer represents a malignancy characterized by high incidence and mortality rates, necessitating further elucidation of its underlying mechanisms. We conducted an analysis using bulk transcriptomic data of ovarian cancer and normal ovarian tissues, as well as single-cell sequencing data according to publicly available databases. Through calculation of Gene Set Variation Analysis (GSVA) scores for TNF family genes, weighted gene co-expression network analysis (WGCNA) for hub genes identification, and subsequent Gene Ontology (GO) enrichment analysis, we delineated pathways crucial in ovarian cancer pathogenesis. Furthermore, differential expression gene analysis facilitated the identification of genes with pronounced expression levels in tumor tissues and their intersection with hub genes, followed by GO analyses across molecular functions (MF), cellular components (CC), and biological processes (BP). Utilizing multivariable Cox regression and LASSO analyses, we constructed a prognostic model comprising 14 genes (GFPT2, PDE4B, PODNL1, TGFBI, CSF1R, PTGIS, SFRP2, COL5A2, TRAC, SLAMF7, VCAN, GBP1P1, C2, TRBV28). Both training and validation sets demonstrated robust diagnostic and prognostic capabilities. Clinical information and immune cell infiltration analyses were further conducted based on the model. In the single-cell sequencing analysis, reducing dimensional complexity and classifying cell types were performed, followed by exploration of gene expression patterns within each subtype and investigation of temporal expression variations across cell subtypes. Biological functional exploration and drug sensitivity analyses were also conducted. Our study contributes novel insights and theoretical foundations for prognosis, treatment, and development of drugs in patients.