Rectal cancer is a malignant tumor with poor prognosis. Identification of prognostic biomarkers is needed to improve overall survival of rectal cancer patients. Here, we firstly identified miR-20a-5p significantly classifying high-risk group and low-risk group of rectal cancer patients. We also found that several known miRNAs miR-142-5p, miR-486-5p, miR-490-3p and miR-133a-3p played important roles in rectal cancer. Secondly, we constructed and analyzed a rectal cancer-related miRNA-mRNA network. A rectal cancer-related functional module was identified from the miRNA-mRNA network. Survival analysis demonstrated great prognosis capacity of the module to distinguish rectal cancer patients. Thirdly, a rectal cancer-related miRNA-lncRNA network was constructed, which followed power law distribution. Hub miRNAs and lncRNAs of the network were suggested to show significant prognosis ability and be enriched in cancer-related pathways. Fourthly, we constructed a rectal cancer-related ceRNA network and detected several typical lncRNA-miRNA-mRNA crosstalk, such as HAND2-AS1, HAND2 and miR-20a-5p crosstalk and MBNL1-AS1, miR-429 and LONRF2 crosstalk, which were validated to function in improving overall survival of rectal cancer patients. Finally, we identified the regulatory feedback that was constituted by transcriptional factors and lncRNAs, including MEIS1, MEIS2 and multiple lncRNAs. We also demonstrated that these lncRNAs were high related to immune cell infiltration. All these results can help us to uncover the molecular mechanism and provide new light on miRNA-mediated gene crosstalks in rectal cancer.

The tumor microenvironment (TME) has been shown to be involved in angiogenesis, tumor metastasis, and immune response, thereby affecting the treatment and prognosis of patients. This study aims to identify genes that are dysregulated in the TME of patients with colon adenocarcinoma (COAD) and to evaluate their prognostic value based on RNA omics data. We obtained 512 COAD samples from the Cancer Genome Atlas (TCGA) database and 579 COAD patients from the independent dataset (GSE39582) in the Gene Expression Omnibus (GEO) database. The immune/stromal/ESTIMATE score of each patient based on their gene expression was calculated using the ESTIMATE algorithm. Kaplan–Meier survival analysis, Cox regression analysis, gene functional enrichment analysis, and protein–protein interaction (PPI) network analysis were performed. We found that immune and stromal scores were significantly correlated with COAD patients’ overall survival (log rank p < 0.05). By comparing the high immune/stromal score group with the low score group, we identified 688 intersection differentially expressed genes (DEGs) from the TCGA dataset (663 upregulated and 25 downregulated). The functional enrichment analysis of intersection DEGs showed that they were mainly enriched in the immune process, cell migration, cell motility, Toll-like receptor signaling pathway, and PI3K-Akt signaling pathway. The hub genes were revealed by PPI network analysis. Through Kaplan–Meier and Cox analysis, four TME-related genes that were significantly related to the prognosis of COAD patients were verified in GSE39582. In addition, we uncovered the relationship between the four prognostic genes and immune cells in COAD. In conclusion, based on the RNA expression profiles of 1091 COAD patients, we screened four genes that can predict prognosis from the TME, which may serve as candidate prognostic biomarkers for COAD.