Fig. 4: Cell type-specific eQTLs of HERVs and their genetic states in immune cells.

a Bar plot showing the number of identified independent eQTLs of HERVs in each cell type. b Upset plot of unique and shared HERV eQTLs across different cell types. c Pie chart showing the proportions of unique and shared eQTLs. d Box plot showing the differential expression of cell-type-specific HERV eQTL examples across different cell types. Upper: CD4-T-specific eQTL between rs6780016 and THE1D_dup769-chr3 expression. Lower: NK cell-specific eQTL between rs28894981 and LTR9_dup12-chr6 expression. Statistical significance was determined using TensorQTL to evaluate allele dosage effects on cell-type-level HERV expression. Cis-regulatory associations were mapped through 10,000 permutations using the map_cis() function, generating empirical p-values and genome-wide false discovery rates. Center lines represent medians, box limits indicate the 25th and 75th percentiles, whiskers extend to 1.5× interquartile range (IQR) from the box edges. rs6780016 n(CC) = 539, n(CT) = 375, n(TT) = 67; rs28894981 n(CC) = 880, n(CT) = 99, n(TT) = 2. e Relative frequency distribution of the distance between eSNPs and the TSS of associated HERVs. f Dot plot showing the enrichment of cell type-specific eSNPs in cell type-specific enhancers. P-value and Odds ratio were calculated using two-sided Fisher’s exact tests comparing the observed overlap of HERVs with each promoter and enhancer region against background expectations. Multiple testing correction was performed using the Benjamini–Hochberg false discovery rate (FDR) method (threshold = 0.05) implemented via the p.adjust function in R. g Genome track of RNA-seq and H3K27ac Chip-seq data around HERV THE1D_dup769-chr3 and LTR9_dup12-chr6, and eSNP rs6780016 and rs28894981 in each cell type.