A Three-Gene Peripheral Blood Potential Diagnosis Signature for Acute Rejection in Renal Transplantation

Abstract

Background: Acute rejection (AR) remains a major issue that negatively impacts long-term allograft survival in renal transplantation. The current study aims to apply machine learning methods to develop a non-invasive diagnostic test for AR based on gene signature in peripheral blood.Methods: We collected blood gene expression profiles of 251 renal transplant patients with biopsy-proven renal status from three independent cohorts in the Gene Expression Omnibus database. After differential expression analysis and machine learning algorithms, selected biomarkers were applied to the least absolute shrinkage and selection operator (LASSO) logistic regression to construct a diagnostic model in the training cohort. The diagnostic ability of the model was further tested in validation cohorts. Gene set enrichment analysis and immune cell assessment were also conducted for further investigation.Results: A novel diagnostic model based on three genes (TSEN15, CAPRIN1 and PRR34-AS1) was constructed in the training cohort (AUC = 0.968) and successfully verified in the validation cohort (AUC = 0.925) with high accuracy. Moreover, the diagnostic model also showed a promising value in discriminating T cell-mediated rejection (TCMR) (AUC = 0.786). Functional enrichment analysis and immune cell evaluation demonstrated that the AR model was significantly correlated with adaptive immunity, especially T cell subsets and dendritic cells.Conclusion: We identified and validated a novel three-gene diagnostic model with high accuracy for AR in renal transplant patients, and the model also performed well in distinguishing TCMR. The current study provided a promising tool to be used as a precise and cost-effective non-invasive test in clinical practice.

Characteristic Genes and Immune Infiltration Analysis for Acute Rejection after Kidney Transplantation

“…In the early stage, AR is still reversible. Early prediction and early prevention of AR are conducive to preventing graft dysfunction and improving the prognosis of patients after transplantation [ 7 ]. Therefore, it is necessary to screen the characteristic genes of AR after renal transplantation.…”

“…The SVM-RFE algorithm has been applied to genomics [ 14 ], which proves its strong performance. In multiple studies, ML has been used to investigate specific biomarkers associated with AR [ 7 , 15 , 16 ]. However, because the current AR study is only through simple difference analysis or a single queue, which cannot guarantee the accuracy of prediction, we intend to find a more accurate AR gene signature through ML technology and multiqueue verification.…”