Fig. 5
Results of polygenic scoring analysis of PAM GWAS and 29 published traits. a Stratified violin plots for PAM → published trait polygenic analyses, showing –log10(p-values) for genetic score (GRS) associations of inferior temporal cortex (IT) PAM and midfrontal cortex (MF) PAM across all p-value thresholds ranging from 0 to 0.5, with 5.0 × 10−5 regular increments (10,000 total scores). Each PAM genetic score was mapped onto the summary statistics of each published GWAS trait and tested for significance. The dotted line represents a corrected statistical significance threshold of p = 8.6 × 10−4, corrected for 29 GWAS traits and two PAM measures. The width of each violin represents the density of PAM polygenic scores associated with each trait at a given significance. For example, a bottom-skewed violin (e.g. educational attainment (Edu)) indicates that a majority of scores across the tested set of thresholds tended to achieve greater significance, whereas a top-skewed violin (e.g. bipolar disorder (BPD)) indicates that a majority of tested scores tended toward lower significance. Peaks achieving at least one score above corrected statistical significance are labeled for their respective GWAS trait. b Network plot illustrating significant results for PAM → GWAS trait analyses, where an arrow indicates a directional effect of a peak PAM GRS on a GWAS trait at corrected significance (edges are labeled with % variance explained in GWAS trait by optimal PAM GRS, and edge thickness is proportional to the –log(p-value) of the association). c Stratified violin plots for the GWAS trait → PAM analyses, such that GRS were first calculated across thresholds for published traits and then tested against the PAM summary statistics. d Network plot illustrating significant results of the GWAS trait → PAM analyses. GRS genetic risk score, AD Alzheimer’s disease, Edu educational attainment, IBDCD irritable bowel disease, Crohn’s disease; IBDUC irritable bowel disease, ulcerative colitis, CAD coronary artery disease, HIPP hippocampal volume (from MRI), RA rheumatoid arthritis, BPD bipolar disorder, MDD major depressive disorder, HOMA homeostasis model assessment (from fasting insulin and glucose); AtopDerm atopic dermatitis, APOB circulating apolipoprotein B, HDL high-density lipoprotein cholesterol, BMDneck bone marrow density of the neck, SCZ schizophrenia, ADHD attention deficit/hyperactivity disorder. The full list of abbreviations found in the legend and descriptions of each published GWAS, including sample sizes, are listed in Supplementary Data 7
