Fig. 1: Two epigenetic biotypes G1 and G2 were identified based on 100 clinical-meaningful DNAm genes.

a Thirty-four well-characterized symptomatic assessments (Table S1) were grouped into five clusters based on Pearson’s correlations values; we assigned clinical-related names to these groups (labels on the left). The values were translated to a color gradient schema in which, red indicates a value of 1 (i.e., positive correlation) and blue indicates −1 (i.e., negative correlation). b Heatmap illustrating the z-scores of the components of two CCA latent clinical factors. The first factor is focused on PTSD core symptoms, named ‘Psychological’, and the second factor contains physical and dissociative, named ‘Physical and dissociative’. Colors span dark blue to dark red where dark blue denotes a z-score of −2, and dark red indicates a z-score of 2. Black boxes were added to highlight the components highly ranked based on absolute z-scores. c The scatter plot shows the DNAm Psychological and Physical score paired with their associated Psychological and Physical clinical features. The highly correlated pairs were identified by PCA + CCA approach based on Discovery cohort (N = 157, r = 0.49, p = 3e−8, and r = 0.41, p = 2.6e−4, respectively). The gray dots are controls, and PTSD individuals are presented in colored dots. d LDA classifier was trained on DNAm ‘Psychological’ and ‘Physical and dissociative’ to assign any PTSD individuals to G1 or G2 biotypes. The gray dots are controls, the blue and red dots are PTSD G1 and G2 biotype.