Mendelian Randomization and Single-Cell RNA Sequencing Reveal CKAP4 and PFDN5 as Tumor Cell-Specific Causal Genes for Glioblastoma.

Background: Glioblastoma (GBM) is an aggressive primary brain tumor with unclear etiology. We aimed to identify causal risk genes by integrating single-cell RNA sequencing (scRNA-seq) and Mendelian randomization (MR). Methods: Differentially expressed genes (DEGs) from public databases were overlapped and analyzed via MR to screen for causal genes. A prognostic model was built using these genes and validated through immune infiltration analysis, single-cell mapping, and experimental assays (RT-qPCR, Western blot). Results: A 6-gene prognostic signature (TMEM158, HOXB2, CKAP4, PEPD, PFDN5, NPC2) was established, where higher risk scores correlated with poorer overall survival and distinct immune profiles. scRNA-seq confirmed tumor cell-specific expression, validated experimentally. Multivariate MR highlighted CKAP4 and PFDN5 as having direct causal links to GBM. Conclusion: The 6-gene signature predicts GBM prognosis, and CKAP4/PFDN5 are promising causal biomarkers and therapeutic targets. This integrated approach provides novel molecular insights and supports personalized therapy development in GBM.