Establishment and verification of a prognostic tumor microenvironment-based and immune-related gene signature in colon cancer

Construction of a prognostic model for breast cancer based on moonlighting genes

Breast cancer (BRCA) is a highly heterogeneous disease, with significant differences in prognosis among patients. Existing biomarkers and prognostic models have limited ability to predict BRCA prognosis. Moonlighting genes regulate tumor progression and are associated with cancer prognosis. This study aimed to construct a moonlighting gene-based prognostic model for BRCA. We obtained differentially expressed genes (DEGs) in BRCA from The Cancer Genome Atlas and intersected them with moonlighting genes from MoonProt to acquire differential moonlighting genes. GO and KEGG results showed main enrichment of these genes in the response of BRCA cells to environmental stimuli and pentose phosphate pathway. Based on moonlighting genes, we conducted drug prediction and validated results through cellular experiments. After ABCB1 knockdown, viability and proliferation of BRCA cells were significantly enhanced. Based on differential moonlighting genes, BRCA was divided into three subgroups, among which cluster2 had the highest survival rate and immunophenoscore and relatively low tumor mutation burden. TP53 had the highest mutation frequency in cluster2 and cluster3, while PIK3CA had a higher mutation frequency in cluster1, with the majority being missense mutations. Subsequently, we established an 11-gene prognostic model in the training set based on DEGs among subgroups using univariate Cox regression, LASSO regression, and multivariable Cox regression analyses. Model prognostic performance was verified in GEO, METABRIC and ICGC validation sets. In summary, this study obtained three BRCA moonlighting gene-related subtypes and constructed an 11-gene prognostic model. The 11-gene BRCA prognostic model has good predictive performance, guiding BRCA prognosis for clinical doctors.

Construction of a prognostic assessment model for colon cancer patients based on immune-related genes and exploration of related immune characteristics

Objectives: To establish a novel risk score model that could predict the survival and immune response of patients with colon cancer. Methods: We used The Cancer Genome Atlas (TCGA) database to get mRNA expression profile data, corresponding clinical information and somatic mutation data of patients with colon cancer. Limma R software package and univariate Cox regression were performed to screen out immune-related prognostic genes. GO (Gene ontology) and KEGG (Kyoto Encyclopedia of Genes and Genomes) were used for gene function enrichment analysis. The risk scoring model was established by Lasso regression and multivariate Cox regression. CIBERSORT was conducted to estimate 22 types of tumor-infiltrating immune cells and immune cell functions in tumors. Correlation analysis was used to demonstrate the relationship between the risk score and immune escape potential. Results: 679 immune-related genes were selected from 7846 differentially expressed genes (DEGs). GO and KEGG analysis found that immune-related DEGs were mainly enriched in immune response, complement activation, cytokine-cytokine receptor interaction and so on. Finally, we established a 3 immune-related genes risk scoring model, which was the accurate independent predictor of overall survival (OS) in colon cancer. Correlation analysis indicated that there were significant differences in T cell exclusion potential in low-risk and high-risk groups. Conclusion: The immune-related gene risk scoring model could contribute to predicting the clinical outcome of patients with colon cancer.

Identification and Verification of m7G Modification Patterns and Characterization of Tumor Microenvironment Infiltration via Multi-Omics Analysis in Clear Cell Renal Cell Carcinoma

The epigenetic modification of tumorigenesis and progression in neoplasm has been demonstrated in recent studies. Nevertheless, the underlying association of N7-methylguanosine (m⁷G) regulation with molecular heterogeneity and tumor microenvironment (TME) in clear cell renal cell carcinoma (ccRCC) remains unknown. We explored the expression profiles and genetic variation features of m⁷G regulators and identified their correlations with patient outcomes in pan-cancer. Three distinct m⁷G modification patterns, including MGCS1, MGCS2, and MGCS3, were further determined and systematically characterized via multi-omics data in ccRCC. Compared with the other two subtypes, patients in MGCS3 exhibited a lower clinical stage/grade and better prognosis. MGCS1 showed the lowest enrichment of metabolic activities. MGCS2 was characterized by the suppression of immunity. We then established and validated a scoring tool named m7Sig, which could predict the prognosis of ccRCC patients. This study revealed that m⁷G modification played a vital role in the formation of the tumor microenvironment in ccRCC. Evaluating the m⁷G modification landscape helps us to raise awareness and strengthen the understanding of ccRCC’s characterization and, furthermore, to guide future clinical decision making.