A four-gene signature for diagnosis of acute kidney injury following kidney transplantation

作者：Jinghan Feng, Xi Zhang, Jiayi Liu, Qian Sun, Fangyuan Peng, Ting Ren, Zhouping Zou, Liu Q, Xiaoqiang Ding, Ping Jia · 发表于：Renal Failure · 年份：2026 · DOI：10.1080/0886022x.2026.2687235 · 研究领域：Acute Kidney Injury Research、Ferroptosis and cancer prognosis、Single-cell and spatial transcriptomics

Acute kidney injury (AKI) is a common and severe complication following renal transplantation, largely driven by ischemia–reperfusion injury (IRI). However, reliable biomarkers for post-transplant AKI remain unavailable. In this study, we analyzed transcriptomic datasets from multiple cohorts to identify robust gene signatures associated with transplant-related AKI. Weighted gene co-expression network and differential expression analyses in the discovery cohort revealed 222 candidate genes altered during early allograft IRI, which were subsequently refined in an integrated training cohort using least absolute shrinkage and selection operator (LASSO) regression and support vector machine–recursive feature elimination (SVM-RFE) to a four-gene signature comprising SOCS3, MYC, TGIF1, and LETM2. This signature showed excellent discriminative performance in the training cohort (10-fold cross-validation AUC = 0.969) and was validated in independent external datasets, including a large cohort (AUC = 0.942). Decision curve analyses indicated potential clinical utility for early AKI identification across a broad threshold-probability range, with favorable performance compared with several established biomarkers such as neutrophil gelatinase–associated lipocalin. Single-cell transcriptomics revealed cell type–specific expression of the four genes across renal compartments. Moreover, their protein levels were elevated at 24 h after IRI in mouse kidneys, and displayed high expression in h...