We identified cis and trans driver TFs underlying epigenetic and transcriptome changes between individuals with and without AUD using a linear regression model, \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$Y=A\beta+{\beta }_{0}+\varepsilon$$\end{document}Y=Aβ+β0+ε. For transcriptome drivers, the regression model predicted the log transformation of the gene expression fold change between individuals with and without AUD (Y) from cell type-specific TF-TG trans-regulation (A). For epigenetic drivers, the model predicted chromatin accessibility changes (Y) from the cell type-specific TF-regulatory element cis-regulation (A). Significant TFs from each model indicated TFs driving differential expression and chromatin states between conditions through direct epigenetic or transcriptome regulation. TFs with (BH corrected) FDR < 0.05 were deemed significant.
