Fig. 7: Construction of a novel fibrosis assessment model to aid BA clinical diagnosis and autologous liver survival prediction. | Pediatric Research

Fig. 7: Construction of a novel fibrosis assessment model to aid BA clinical diagnosis and autologous liver survival prediction.

From: Scar-associated macrophages and biliary epithelial cells interaction exacerbates hepatic fibrosis in biliary atresia

Fig. 7

a Venn diagram showing overlapping genes identified by SPARK as highly variable, highly expressed in fibrotic regions, spatial BA-related genes, and single-cell BA highly expressed genes. b Integration of 15 machine learning approaches for hub genes screening of overlapping genes for BA diagnosis. c–e ROC curves demonstrating the diagnostic efficacy of four genetic models for BA in GSE46960 (train set), GSE122340 (test set), and GSE221346 (test set). f Four genes classify GSE15235 into two subgroups, high and low expression. g Differences in autologous liver survival in two subpopulations of GSE15235. BA biliary atresia, ROC Receiver operating characteristic curve.

Back to article page