Heterogeneous epigenetic variation converges on splicing dysregulation in opioid addiction

Disease heterogeneity presents a major challenge for genetic and epigenetic dissection of complex traits. Neuropsychiatric traits, such as opioid use disorder (OUD), arise from diverse genetic and environmental factors that uniquely combine in individuals, complicating efforts to identify causal genes and enhancers. We identified epigenetic variation linked to opioid overdose by profiling H3K27ac in the nucleus accumbens (NAc) of 91 cases and controls. While standard approaches identified only 156 loci with differential acetylation, machine learning models built on combinations of putative regulatory regions distinguished cases from controls at high accuracy (>0.95) and were significantly enriched for OUD genetic risk. To leverage disease heterogeneity, we defined individual-specific variation in H3K27ac, termed Variant Enhancer Loci (VEL). VELs converged on shared target genes at a five-fold higher rate than prior prefrontal cortex studies. RNA splicing was the top-enriched ontology, including the splicing regulator CELF5. CELF5 knockdown in iPSC-derived medium spiny neurons was linked to collagen proteins and extracellular matrix organization. Collectively, this study provides new insights into opioid addiction and a framework for personalized epigenetic analysis of heterogeneous diseases.