AUTHOR=Yin Yu , Chen Congcong , Zhang Dong , Han Qianguang , Wang Zijie , Huang Zhengkai , Chen Hao , Sun Li , Fei Shuang , Tao Jun , Han Zhijian , Tan Ruoyun , Gu Min , Ju Xiaobing 

TITLE=Construction of predictive model of interstitial fibrosis and tubular atrophy after kidney transplantation with machine learning algorithms

JOURNAL=Frontiers in Genetics

VOLUME=Volume 14 - 2023

YEAR=2023

URL=https://www.frontiersin.org/journals/genetics/articles/10.3389/fgene.2023.1276963

DOI=10.3389/fgene.2023.1276963

ISSN=1664-8021

ABSTRACT=Background: Interstitial fibrosis and tubular atrophy (IFTA) are the histopathological manifestations of CKD and one of the causes of long-term renal loss in transplanted kidneys. Necroptosis as a type of programmed death is an important role in the development of IFTA as well as in the late functional decline and even in the loss of grafts. In this study, 13 machine learning algorithms were used to construct IFTA diagnostic models based on necroptosis-related genes. Methods: We screened all 162 "kidney transplant" related cohorts in the GEO database and obtained 5 data sets (training sets: GSE98320, GSE76882 validation sets: GSE22459, GSE53605 survival sets: GSE21374). Train set was constructed after removing batch effects of GSE98320, GSE76882 by sva package. Differentially expressed genes (DEGs) analysis was used to identify necroptosisrelated DEGs. A total of 13 machine learning algorithms including Lasso, Ridge, Enet, Stepglm, SVM, glmBoost, LDA, plsRglm, RandomForest, GBM, XGBoost, NaiveBayes, ANNs were used to construct 114 IFTA diagnostic models and the optimal models were screened by the AUC values. Post-transplant patients were then grouped using consensus clustering, and the different subgroups were further explored using PCA, KM Survival Analysis, Functional Enrichment Analysis, CIBERSOFT, and single-sample Gene Set Enrichment Analysis. Results: A total of 55 necroptosis-related DEGs were identified by taking the intersection of the DEGs and necroptosis-related gene sets. Stepglm[both]+RF is the optimal model with an average AUC of 0.822. A total of four molecular subgroups of renal transplant patients were obtained by clustering, and significant upregulation of fibrosis-related pathways as well as upregulation of immune-response-related pathways were found in the C4 group, which had a poorer prognosis.Basing on the combination of 13 machine learning algorithms, we developed 114 IFTA classification models. Furthermore, we test the top model using 2 independent datasets from GEO.