Multi-Omics Analysis of Aberrances and Functional Implications of IRF5 in Digestive Tract Tumours

Oesophageal cancer (EC) is a common gastrointestinal malignancy and includes oesophageal squamous cell carcinoma (ESCC) and oesophageal adenocarcinoma (EAC) sub-types. Gene signatures predicting patient outcomes are not routinely used in clinical practice, particularly owing to batch effects and data standardisation. Here, we sought to establish and validate a reliable signature of senescence-related genes (SRGs) that would aid in predicting prognosis in patients with EC. We downloaded transcriptomics data, and a novel pairwise comparison algorithm selected valid SRG pairs (SRGPs) to construct a prognostic SRGP signature. The SRGPs were verified using Kaplan-Meier survival and receiver operating characteristic curve analyses. Additionally, the relationships between the SRGP signatures and prognosis, immune cell infiltration and chemotherapeutic drug responsiveness were evaluated. The random forest algorithm identified the most clinically significant genes, followed by experimental validation. 19 and 26 SRGP signatures were created for ESCC (n = 81) and EAC (n = 79), respectively. Patients with EC were divided into two groups based on the median risk score. The Kaplan-Meier analysis demonstrated significant differences in overall survival between the ESCC and EAC groups (p < 0.001). The sub-types exhibited different immune signatures. IRF5 was the most clinically significant gene for ESCC. It was highly expressed in ESCC cells, and IRF5 knockdown inhibited cell migration and proliferation, while promoting apoptosis and senescence. The SRGP signature may predict prognosis and immunotherapeutic responses, and IRF5 is a potential target gene for ESCC.