Deciphering the prognostic significance of liquid–liquid phase separation-related genes in thyroid carcinoma: A transcriptomic analysis

作者：Wenjun Zhao, Peng Ma, Yuesheng Wu, Kai Ma, Lei Zhang, Jian Wang, Jun Zhao · 发表于：Cancer Plus · 年份：2026 · DOI：10.36922/cp025320050

Thyroid carcinoma (THCA) is a common endocrine malignancy with heterogeneous clinical outcomes, and the prognostic significance of liquid–liquid phase separation (LLPS)-related genes in THCA remains unexplored. This study focuses on THCA, aiming to explore LLPS-related prognostic genes to provide novel theoretical foundations for its diagnosis and treatment. Utilizing public datasets, we identified LLPS-associated prognostic genes in THCA via differential expression analysis, Cox regression, and machine learning. A risk prediction model was subsequently constructed, followed by drug-sensitivity analysis, while the expression of prognostic genes was validated using reverse transcription-quantitative polymerase chain reaction (RT-qPCR). The results revealed seven key prognostic genes: APOE, ARHGEF4, CCL17, DYRK3, EGF, HSPA1A, and MAP1B. The risk model demonstrated strong predictive performance, with area under the curve values exceeding 0.7 for survival rate prediction. Cox regression analysis indicated that both the risk score and age were significant prognostic factors, with the risk score exhibiting a stronger independent prognostic effect (hazard ratio = 2.328). The nomogram accurately predicted survival, and drug-sensitivity analysis showed significant differences in responses to paclitaxel and cyclophosphamide between different risk groups. RT-qPCR validation confirmed upregulation of APOE, CCL17, ARHGEF4, DYRK3, HSPA1A, and MAP1B, along with downregulation of EGF in the ...