Fig. 4

a Manhattan plot showing the \(-\log _{10}\) p value of association (asymptotic PERMANOVA test as implemented in MANTA) between genetic variants and the volumes of 12 hippocampal subfields. The horizontal black line corresponds to the genome-wide significance threshold selected (5 \(\cdot\) 10\(^{-8}\)). b Regional plot of the GMNC locus, showing the \(-\log _{10}\) p values for association with hippocampal subfield volumes for all tested genetic variants (asymptotic PERMANOVA test). Linkage disequilibrium patterns (color-coded) and recombination rates are also displayed. The lower panel represents the location of previous associations with brain-related traits and diseases in the GWAS Catalog (shown as arrows), encoded as follows: (1) brain morphology, (2) subcortical volume, (3) nucleus accumbens volume, (4) cortical thickness, (5) lateral ventricular volume, (6) brain region volumes, (7) cortical surface area, (8) white matter microstructure, (9) cerebrospinal P-tau181p levels, (10) cerebrospinal T-tau levels, and (11) Alzheimer’s disease biomarkers. c Covariate-adjusted volumes (mm\(^3\)) of the most changing hippocampal subfields across genotype groups at rs9877502 (chr3:190,669,518, G/A) are shown as boxplots. In boxplots, the box represents the first to third quartiles and the median, while the whiskers indicate ± 1.5 \(\times\) interquartile range. The number of individuals on each group is shown between parentheses. FDR adjusted p values (Wilcoxon Rank-Sum test) for each pairwise comparison are also displayed, encoded as follows: *** (p \(\le\) 0.001), ** (0.001 < p \(\le\) 0.01), * (0.01 < p \(\le\) 0.05), n.s. (p > 0.05)