Using bioinformatics methods to elucidate fatty acid-binding protein 4 as a potential biomarker for colon adenocarcinoma

BACKGROUND Colon adenocarcinoma (COAD) ranks second in terms of cancer-related deaths. We found that fatty acid-binding protein 4 (FABP4), which is related to cell adhesion and immunity, affects the occurrence and development of COAD. This study focused on the possibility of using FABP4 as a biomarker for COAD and constructed a nomogram for predicting the survival of COAD patients. AIM To verify the possibility of using FABP4 as a biomarker for COAD. METHODS A total of 453 COAD tissue samples, along with 41 normal tissue samples, were obtained from The Cancer Genome Atlas database. The difference in FABP4 expression between COAD tissues and normal tissues was analyzed, and the results were verified by immunohistochemistry. The WGCNA algorithm links FABP4 expression with an enrichment analysis and with immune cell infiltration pathways. The biological functions of FABP4 and its coexpressed genes were explored through enrichment analyses. The ESTIMATE, CIBERSORT and ssGSEA methods were used for the immune infiltration analysis. Finally, risk scores were calculated by a Cox analysis. A nomogram was constructed by combining risk scores with routine clinicopathological factors. We assessed the accuracy of survival predictions based on the C-index. The C-index ranges from 0.5 to 1.0, and in general, a C-index value greater than 0.65 indicates a reasonable estimate. The results were validated using the Gene Expression Omnibus (GEO) database. RESULTS FABP4 was significantly differentially expressed in COAD. It is a promising auxiliary biomarker for screening and diagnosis. Enrichment analyses suggested that FABP4 may influence the invasion and progression of COAD through cell adhesion. The immunological analysis revealed that FABP4 expression in COAD was significantly positively correlated with immune cell infiltration. Moreover, a nomogram to predict the survival of COAD patients was successfully constructed by integrating the calculated risk scores of 15 candidate genes and routine clinicopathological factors. This nomogram could effectively predict 1-year, 3-year, and 5-year survival (C-index = 0.786) and was verified (C-index = 0.73). CONCLUSION This study established FABP4 as an effective biomarker for screening, assisting in the diagnosis and determining the prognosis.