Fig. 3: Sex-specific rsFC changes in five transcriptomic regions of interest.

a–c Color maps of depressed vs healthy t-statistics showing the neuro-anatomical distribution of rsFC abnormalities that are (a) female-specific, b shared between sexes, or (c) male-specific in five transcriptomic seed ROIs labeled in rows, depicting the peak t-statistic across both hemispheres for each functional parcel. Parcels with significant (FDR q < 0.05) main effects of MDD or MDD-by-sex interactions are outlined in green and presented in detail in Supplementary Fig. S4. To illustrate how sex-specificity is not driven by stringent thresholding, we also plot nominally significant effects (unadjusted p < 0.05) in the parcels not outlined in green. d Venn diagrams depicting the number of sex-specific vs. shared connectivity effects at a liberal threshold (unadjusted p < 0.05) at each seed ROI. Sex specificity was even more pronounced at more stringent thresholds (Supplementary Fig. S4c, d). e Boxplots depicting the distribution of female- (left) and male-specific (right) connectivity effects in a null model with sex labels randomly permuted 1000 times. Red dots represent the empirically observed number of sex-specific effects from (d). † = p < 0.1; * = p < 0.05; ** = p < 0.01.