Fig. 3: Enhancer usage near hundreds of genes associates with seizure severity.

A Representative genome browser track depicting number of insertions, insertion density, peaks annotated by CC pipeline, and H3K27ac ChIP-seq density from P0 mouse forebrain (ENCODE) around the Snx10 and Skap2 loci (chr6: 51500K–51900K). Insertions enriched in mild or severe responses are separated, and annotated peaks represent genomic locations with strong density across responders. The H3K27ac track demonstrates a presence of established enhancer sites near peaks. B Enrichment plot (top) and heatmap (bottom) of H3K27ac ChIP-seq (same as A) relative enrichment, quantified as fold change over ChIP-seq control, centered on CC peaks. C Volcano plot depicting log2 fold enrichment (x-axis) and adjusted p-values (y-axis) of differential peaks between mild (teal) and severe responders (purple) using Fisher’s exact test with Benjamini–Hochberg correction. Peaks are annotated with nearby genes. D Genome browser track for representative peak around Ramac/Homer2 (chr7:81775773-81779508) depicting differential enrichment of mild insertions. E Genome browser track for multiple representative peaks around Zic1/Zic4 (chr9:91355K-91450K) depicting differential enrichment of severe insertions. F–I Two-sided Spearman correlations between individual animal seizure scores and insertion counts at loci near (F) Gabrb1/Gabra4 (Spearman ρ = −0.73, p = 0.02), G Ramac/Homer2 (ρ = −0.71, p = 0.02), H Man1a (ρ = 0.81, p = 0.004), and I Kcnc1 (ρ = 0.72, p = 0.02).