Linking regulatory variants to target genes by integrating single-cell multiome methods and genomic distance.

Methods that analyze single-cell paired RNA sequencing (RNA-seq) and assay for transposase-accessible chromatin using sequencing (ATAC-seq) multiome data have shown promise in linking regulatory elements to genes. However, existing methods exhibit low concordance and do not capture the effects of genomic distance. We propose pgBoost, an integrative modeling framework that trains a non-linear combination of existing linking strategies (including genomic distance) on expression quantitative trait locus (eQTL) data to assign a probabilistic score to each candidate single-nucleotide polymorphism-gene link. pgBoost attained higher enrichment than existing methods for evaluation sets derived from eQTL, activity-by-contact, CRISPR and genome-wide association study (GWAS) data. We further determined that restricting pgBoost to features from a focal cell type improved power to identify links relevant to that cell type. We highlight several examples in which pgBoost linked fine-mapped GWAS variants to experimentally validated or biologically plausible target genes that were not implicated by other methods. In conclusion, a non-linear combination of linking strategies improves power to identify target genes underlying GWAS associations.