We aimed to construct and validate a prognostic-predicting model of rectum adenocarcinoma (READ) based on RNA-binding protein-related genes (RBPGs) by bioinformatics and statistical analysis. We obtained the expression matrix containing 1542 RBPGs from the RBPDB database through the
R package. Then, 126 differentially expressed RBPGs (DE-RBPGs) were obtained by differential expression analysis between groups, among which 63 down-regulated genes and 63 up-regulated genes. Next, Ribonucleoprotein complex biosynthesis and assembly were the primary biological processes (BP)
identified by the Gene Ontology (GO) enrichment study, cytoplasmic translation, ncRNA processing, ncRNA and rRNA metabolic process. The functions of cellular components (CC) were closely related to organellar and mitochondrial ribosomes and their subunits, spliceosomal complex, mitochondrial
matrix and ribonucleoprotein granule. Then, we put 126 DE-RBPGs into the protein–protein interaction (PPI) network, showing the mutual regulation between each DE-RBPGs. In addition, eight prognostics DE-RBPGs (PDE-RBPGs) were identified by Cox regression analysis, among which DIS3L,
EFTUD2, FAM98B, IREB2, NOP58, PDCD7 and STRBP were low-risk PDE-RBPGs (HR less than 1), while GTF3A was a high-risk PDE-RBPG (HR greater than 1). A prognosis model consisting of two PDE-RBPGs (EFTUD2 and FAM98B) was finally optimized. The results of the study
of the Receiver Operating Characteristics (ROC) curve and the survival analysis revealed that the prognostic-predicting model constructed by us could accurately predict the grouping and prognosis of READ patients. The above results further elucidated the important molecular functions, key
biological pathways and gene (protein) interactions of DE-RBPGs. The prognostic-predicting model constructed by us can accurately predict the patients with READ, which is very valuable as a guide for READ early clinical evaluation and therapy.