Integrative profiling of diverse post-translational modifications for prognostic stratification and personalized therapy in papillary thyroid cancer

Background: Post-translational modifications (PTMs) play critical roles in the pathogenesis and progression of tumors. However, the biological and clinical relevance of PTMs in papillary thyroid cancer (PTC) remains unexplored.Methods: We collected bulk transcriptome, single-cell transcriptome, and pertinent clinicopathological details of PTC patients from The Cancer Genome Atlas (TCGA) and Gene Expression Omnibus (GEO) databases. We analyzed seventeen PTMs and integrated 12 distinct machine learning algorithms to construct a prognostic model for predicting the thyroid cancer-free interval (TCFi). Enrichment analysis, single-cell analysis, and immune-related analysis were performed to elucidate the biological role of PTMs. Therapeutic responses of PTC patients were predicted based on the model. The expression of model genes was further validated, and we identified the key signature associated with the malignant phenotypes of PTC.Findings: We filtered out twelve genes to construct a post-translational modification index (PTMI) and identified three molecular clusters of PTC by unsupervised clustering based on model genes. Shapley additive explanations (SHAP) and nomogram models provide additional confirmation of the predictive efficacy of PTMI. Integrated analyses revealed significant associations between PTMI and immune features. Additionally, the high-PTMI group may be more responsive to FDA-supported drugs and conventional chemotherapeutics, while exhibiting resistance to radioactive iodine (RAI) therapy. Notably, we identified TYMS as a key functional PTMI signature.Interpretation: The PTMI proposed in this study holds strong potential as a prognostic biomarker and therapeutic predictor, offering valuable insights for personalized management of PTC patients.