Fig. 1.

The workflow of current integrative genomics analysis. a Integrating single cell RNA-sequencing data with GWAS summary statistics on PBC based on a regression-based polygenic model; Left panel: Two independent single cell RNA-seq datasets based on liver and peripheral CD4 + T cells, and one large-scale GWAS summary statistics on PBC; Middle panel: There were 8444 parenchymal and non-parenchymal cells obtained the transcriptional profiles based on the CellRanger analysis pipeline and clustered by using t-Distributed Stochastic Neighbor Embedding (tSNE) method, and we also obtained 1,124,241 SNPs from the whole genome with P values and Beta value; Right panel: We applied a regression-based polygenic model in RolyPoly to unveil the genetic mapping single-cell landscape for PBC. b Prioritization of PBC-risk genes by integrating GWAS summary data with GTEx eQTL data; In this step, multiple bioinformatics analyses were leveraged to prioritize PBC-risk genes, including MAMGA-based gene-level association analysis, gene-property analysis, S-MultiXcan- and S-PrediXcan-based integrative genomics analyses, 10⁵ times of in silico permutation analysis, differential gene expression (DGE) analysis, multidimensional scaling (MDS) analysis, PPI network-based analysis, drug-gene interaction analysis, and functional enrichment analysis. c Genetics-influenced liver cell subpopulations and its immune microenvironment for PBC. We performed comprehensive single cell sequencing-based analyses to uncover the biological functions of ORMDL3⁺ cholangiocytes and its interacted immune cells of macrophages and monocytes. ORMDL3⁺ cholangiocytes have significantly elevated metabolism activity score, and VEGF signaling pathway has a crucial role in cellular communications of ORMDL3⁺ cholangiocytes