Acute rejection (AR) is a common complication in the early stage after kidney transplantation. Some studies have shown that the occurrence of AR after kidney transplantation may further affect the development of tumors, and both AR and tumor development are related to immune cells and immune genes, so it is particularly important to diagnose the occurrence of AR at an early stage and to analyze the correlation between AR and tumors. In this study, we applied bioinformatics techniques for differential expression analysis and weighted gene co-expression network analysis analysis of AR patients to obtain differentially expressed genes and modular genes significantly associated with AR, respectively, so as to obtain their intersecting genes with immune-related genes; 21 intersecting genes were screened by lasso regression and Boruta algorithm to obtain the genes, and finally, the feature genes that were significantly associated with the dependent variable were further obtained by single-factor and multi-factor logistic regression. Then the best diagnostic model for AR was screened by 10 machine learning methods, and we evaluated the model in various aspects, such as receiver operator characteristic curve, decision curve analysis. We then focused on the role of FAM3C in renal cancer. We finally screened 4 feature genes (CD1D, FPR2, FAM3C, and HMOX1) to construct the AR diagnostic model; through comparative evaluation, we believe that logistic regression shows a better advantage in the construction of diagnostic models for AR. FAM3C may become a potential biological marker for AR diagnosis and plays an important role in the development of renal cancer. In summary, immune-related genes play an important role in the diagnosis of AR after kidney transplantation, and the gene FAM3C may be a potential therapeutic target for AR and renal cancer.
Read full abstract