Biliary Atresia Subtypes Based on the Immune Cells and Genes Expressions by Unsupervised Clustering Imply the Underlying Pathogenesis

Background: The pathogenesis and phenotyping of biliary atresia (BA) is not clear. We aimed to provide a new view to identify the subtype of BA and explore the underlying pathogenesis. Methods: We collected the BA microarray datasets from Gene Expression Omnibus (GEO), and performed an unsupervised cluster analysis on abundance of immune cells and gene expressions and identified de novo classified BA patients. Then, we explored the possible molecular mechanisms of each subtype by functional enrichment analysis and the hub genes. Findings: First we classified BA patients into three subtypes and validated in GSE15235 dataset. Then we identified representative genes and immune cells for each subtype. The result of functional enrichment analysis and hub genes in subtype 2 was correlated with immune receptor activity, cytokine receptor, signaling by interleukin including IL-4, IL-13, IL-10, viral protein interaction, this may infer subtype 2 was corresponded to cytomegalovirus-associated variants of the clinical phenotyping classifies and modulation of interleukin signaling may represent potential novel therapeutic approaches. Neutrophil degranulation by Reactome pathway analysis existed in all the subtypes. Interpretation: The subtypes and function analysis identified in this study help us understand the underlying pathogenesis of BA, and provide new ideas for basic and clinical study. Funding: This work was supported by a grant from the National Key R&D Program of China (2018YFC1315400) and 5010 Project of Sun Yat-sen University (No.2018024). Declaration of Interest: None to declare.