Local SNP-explained methylation variation reveals genetically anchored and exposure-associated methylation...

Human brain DNA methylation is shaped by inherited genetic variation and cumulative environmental experience, yet how these influences partition the methylome remains poorly resolved in postmortem cohorts with modest sample sizes and limited ancestral diversity. To map this architecture in an underrepresented population, we analyzed whole-genome bisulfite sequencing and genotype array data from 168 admixed Black American adults from the BrainSEQ consortium across three brain regions. We adapted and benchmarked SNP-based elastic-net modeling to classify variably methylated regions (VMRs) by local SNP-explained methylation variation, an approach that provided stable classification at the modest sample sizes of postmortem brain cohorts, where conventional methods are underpowered. Using this framework, we partitioned 31,143 VMRs into high and low SNP-explained classes and evaluated their generalizability in a multi-ancestry cohort of Black American and non-Hispanic white American donors. High SNP-explained VMRs were concentrated in distal intergenic sequences and, at the highest heritability levels, were enriched for H3K9me3, quiescent/repressive chromatin states and LINE/L1 elements, linking genetically anchored methylation to repeat-associated repressive chromatin across the human brain. A small subset overlapping Activity-by-Contact-defined enhancers was linked to candidate immune-related genes, including MHC class II loci. By contrast, low SNP-explained VMRs were more gene-proximal and enriched for active regulatory elements. Metadata-associated VMRs showed region-, exposure-, and donor-group-dependent enrichment across SNP-explained classes, including substance use and sociodemographic variables. Together, these findings show that the most genetically anchored component of the human brain methylome is concentrated in repressive, repeat-rich chromatin compartments involved in heterochromatin maintenance and repeat silencing, distinct from the gene-proximal, exposure-associated variation less explained by nearby SNPs. By resolving this architecture in an underrepresented population, this work clarifies how inherited variation structures the brain methylome and, given the established role of these compartments in neuronal aging, informs the interpretation of epigenomic mechanisms relevant to neuropsychiatric and neurodegenerative diseases.