Tissue-specific enhancer-gene maps from multimodal single-cell data identify causal disease alleles

Area of Science:

• Genomics and Bioinformatics
• Molecular Biology
• Computational Biology

Background:

• Translating genome-wide association study (GWAS) findings into causal variants and genes is crucial for understanding disease mechanisms.
• Accurate cell-type-specific enhancer-gene maps are essential but difficult to generate from primary human tissues using existing experimental techniques.

Purpose of the Study:

• To develop a novel computational method for creating cell-type-specific enhancer-gene maps from multimodal single-cell sequencing data.
• To enable the scalable and accurate identification of causal variants and genes associated with human diseases and traits.

Main Methods:

• Developed SCENT (single-cell enhancer target gene mapping), a nonparametric statistical approach.
• Modeled the association between enhancer chromatin accessibility and gene expression using single-cell/nucleus multimodal RNA sequencing and ATAC sequencing data.
• Applied SCENT to 9 multimodal datasets comprising over 120,000 cells/nuclei.

Main Results:

• Generated 23 cell-type-specific enhancer-gene maps.
• These maps showed significant enrichment for causal variants in expression quantitative loci (eQTLs) and GWAS for 1,143 diseases and traits.
• Identified likely causal genes for common and rare diseases and linked somatic mutation hotspots to target genes.

Conclusions:

• SCENT provides a scalable and accurate method for constructing cell-type-specific enhancer-gene maps from disease-relevant human tissues.
• These maps are vital for deciphering the functional impact of noncoding genetic variants.
• The approach facilitates the identification of disease-associated genes and variants at a high resolution.