Integrative Multi-omics and Causal Mediation Analyses Identify an Environmental Exposure-Related Immunoregulatory Axis in Primary Sclerosing Cholangitis
Description
Abstract / Description: This dataset supports the findings of the manuscript entitled “Integrative Multi-omics and Causal Mediation Analyses Identify an Environmental Exposure-Related Immunoregulatory Axis in Primary Sclerosing Cholangitis”. The data include: NHANES (2015–2016) urinary volatile organic compound (VOC) metabolite levels and liver disease phenotype data used for univariable and multivariable logistic regression analyses. Lists of putative 2HBeMA-associated targets retrieved from ChEMBL, STITCH, and SwissTargetPrediction databases (n = 1,314). Lists of primary sclerosing cholangitis (PSC)-related genes retrieved from GeneCards and OMIM databases (n = 3,589). Overlapping gene sets (428 and 58 candidate genes) from intersection analyses. Transcriptomic data from GEO datasets (accession numbers provided in the manuscript) used for differential expression analysis. Bayesian mediation analysis results: significant mediators (CSV file), including Path A, Path B, mediation effect, confidence intervals, and proportion mediated. Molecular docking output (AutoDock Vina) for 2HBeMA with PRKCB: binding affinities and RMSD values for 20 docking modes. Single-cell and spatial transcriptomics processed data used for immune cell communication analysis (CellChat). All data are provided in raw or processed formats as appropriate. Detailed methods are described in the associated manuscript. Keywords: Primary sclerosing cholangitis; volatile organic compounds; 2HBeMA; multi-omics; Mendelian randomization; mediation analysis; molecular docking; NHANES; GEO Usage Notes / Methods: NHANES data were analyzed using R (version 4.2) with survey-weighted logistic regression. Target prediction and gene set intersections were performed using custom Python scripts. Bayesian mediation analysis was conducted using the bayesmed R package (or equivalent). Molecular docking was performed with AutoDock Vina (Trott and Olson, 2010). Single-cell data integration and cell-cell communication inference were performed using Seurat (v4) and CellChat (Jin et al., 2021).