基于生物信息学构建和验证甲状腺乳头状癌中与 parthanatos 相关的预后基因风险模型

Using the GSE33630 dataset, differentially expressed PRGs were identified and analyzed via weighted gene co-expression network analysis (WGCNA) to pinpoint key module genes. Regression analysis selected seven prognostic genes for risk model construction. The model's performance was validated, and a nomogram was developed for survival prediction. Further analyses included clinical feature correlations, immune infiltration, drug sensitivity, gene set enrichment analysis (GSEA), and experimental validation via RT-qPCR.

Seven prognostic genes (TSHZ3, SERGEF, AKAP12, SGPP2, ASGR1, AK1, PELI2) were identified. The risk model demonstrated robust predictive accuracy, stratifying patients into high- and low-risk groups with significant survival differences. GSEA revealed 29 enriched pathways (e.g., ribosome, focal adhesion), while immune infiltration analysis highlighted CD56 + NK cells and AK1 as key immune correlates. Drug sensitivity screening identified 111 differential therapeutics. Functional analysis indicated AKAP12 had the strongest functional similarity among prognostic genes.

This study comprehensively mapped PRGs in PTC, established a validated risk model, and provided insights into immune-microenvironment interactions and therapeutic targets, advancing precision oncology for PTC.