Hierarchical annotation of eQTLs enables identification of genes with cell-type divergent regulation

While context-type-specific regulation of genes is largely determined by cis-regulatory regions, attempts to identify cell-type specific eQTLs are complicated by the nested nature of cell types. We present a network-based model for hierarchical annotation of bulk-derived eQTLs to levels of a cell type tree using single cell chromatin accessibility data and no clustering of cells into discrete cell types. Using our model, we annotated bulk-derived eQTLs from the developing brain with high specificity to levels of a cell-type hierarchy. The increased annotation power provided by the hierarchical model allowed for sensitive detection of genes with multiple distinct non-coding elements regulating their expression in different cell types, which we validated in single-cell multiome data and reporter assays. Overall, we find that incorporating the hierarchical organization of cell types provides a powerful way to account for the relationships between cell types in complex tissues.