Characterization of piwi-RNA signatures in colorectal cancer.

227 Background: Although incidence and mortality rates have declined, colorectal cancer (CRC) remains a leading cause of cancer-related deaths worldwide. Identifying novel diagnostic and prognostic biomarkers, as well as therapeutic targets, is critical for improving patient outcomes. This study evaluates PIWI-interacting RNAs (piRNAs) as potential biomarkers for CRC. Methods: Colorectal Cancer small RNA sequencing data from the Gene Expression Omnibus (GEO) were analyzed to identify differentially expressed piRNAs. Diagnostic models were constructed using multiple machine learning algorithms, with validation performed in The Cancer Genome Atlas Colon Adenocarcinoma (TCGA-COAD) dataset. Prognostic modeling was conducted with LASSO Cox regression in TCGA-COAD and benchmarked against 109 published prognostic signatures. Predicted target genes of candidate piRNAs were further analyzed for pathway enrichment. Results: Among four machine learning classifiers, the support vector machine achieved the best diagnostic performance (AUC = 0.932). A 10-piRNA prognostic signature was significantly associated with poor survival [p < 0.001, HR = 2.76 (1.68–4.5)], with risk further increased in advanced disease [Stage III: p < 0.05, HR = 3.22 (1.12–9.2); Stage IV: p < 0.05, HR = 8.28 (2.88–23.8)]. The 10-piRNA signature retained prognostic significance in younger patients (40–60 years) and those with early-stage disease (Stage I–II). When applied to TCGA-COAD, this model outperformed 109 previously published signatures. Notably, piR-36011 was identified in both diagnostic and prognostic models, with predicted targets involved in oncogenic pathways such as Netrin-1 signaling, MAPK, BRAF, and RAF1. Conclusions: Differentially expressed piRNAs show strong diagnostic and prognostic utility in CRC. The piRNA-based prognostic model outperformed existing signatures, suggesting that piRNAs may represent more reliable biomarkers. In particular, piR-36011 emerges as a promising dual-purpose biomarker linked to colorectal cancer oncogenesis.