A Computational Recognition Analysis of Promising Prognostic Biomarkers in Breast, Colon and Lung Cancer Patients

Breast, colon, and lung carcinomas are classified as aggressive tumors with poor relapse-free survival (RFS), progression-free survival (PF), and poor hazard ratios (HRs) despite extensive therapy. Therefore, it is essential to identify a gene expression signature that correlates with RFS/PF and HR status in order to predict treatment efficiency. RNA-binding proteins (RBPs) play critical roles in RNA metabolism, including RNA transcription, maturation, and post-translational regulation. However, their involvement in cancer is not yet fully understood. In this study, we used computational bioinformatics to classify the functions and correlations of RBPs in solid cancers. We aimed to identify molecular biomarkers that could help predict disease prognosis and improve the therapeutic efficiency in treated patients. Intersection analysis summarized more than 1659 RBPs across three recently updated RNA databases. Bioinformatics analysis showed that 58 RBPs were common in breast, colon, and lung cancers, with HR values < 1 and >1 and a significant Q-value < 0.0001. RBP gene clusters were identified based on RFS/PF, HR, p-value, and fold induction. To define union RBPs, common genes were subjected to hierarchical clustering and were classified into two groups. Poor survival was associated with high genes expression, including CDKN2A, MEX3A, RPL39L, VARS, GSPT1, SNRPE, SSR1, and TIA1 in breast and colon cancer but not with lung cancer; and poor survival was associated with low genes expression, including PPARGC1B, EIF4E3, and SMAD9 in breast, colon, and lung cancer. This study highlights the significant contribution of PPARGC1B, EIF4E3, and SMAD9 out of 11 RBP genes as prognostic predictors in patients with breast, colon, and lung cancers and their potential application in personalized therapy.