Additional file 1: Figure S1. of Genome-wide imaging association study implicates functional activity and glial homeostasis of the caudate in smoking addiction

Resting state functional connectivity (RSFC) among 12 brain regions. Rows and columns indicate the regions that comprise pairs for which RSFCs were computed. White, dark-gray, and light-gray bars represent average z-scores from magnetic resonance imaging measurements for non-smokers, sated smokers, and abstinent smokers, respectively, with standard errors marked above. All tickmarks are spaced 0.1 apart. Asterisks highlight differences that are significant at the 0.01 level. ACC, anterior cingulate cortex; GP, globus pallidus; NAcc, nucleus accumbens; iPFC, inferior prefrontal cortex; mPFC, medial prefrontal cortex; sPFC, superior prefrontal cortex; SMA, supplementary motor area. Figure S2. Quantile-quantile plots of GWAS results. With adjustment for smoker status and population stratification, allelic dosages of germline variants were linearly regressed on (A) caudate activity during disappointment, and (B) putamen activity during reward. The negative logarithms (base 10) of observed P-values were plotted in relation to those of expected P-values. Figure S3. Abstinence-associated withdrawal severity versus caudate activity during reward (CR) and putamen activity during disappointment (PD). After smokers had refrained from smoking for at least 12 h, the severity of their withdrawal symptoms was assessed using the Shiffman-Jarvik Withdrawal Questionnaire (SJWQ) and plotted in relation to (A) CR and (B) PD. As in Table 1, CR and PD are presented as z-scores following normalization that also took into account values from sated smokers and non-smokers. Table S1. Resting state functional connectivity (RSFC) differences among non-smokers, sated smokers, and abstinent smokers. Comparisons were performed using unpaired t-tests. ACC, anterior cingulate cortex; GP, globus pallidus; NAcc, nucleus accumbens; iPFC, inferior prefrontal cortex; mPFC, medial prefrontal cortex; sPFC, superior prefrontal cortex; SMA, supplementary motor area. Table S2. Top 5000 results from GWAS of caudate activity during reward. With adjustment for smoker status and population stratification, linear regression using allelic dosages of germline variants was performed. b, effect size from linear regression; MAF, minor allele frequency; s.e., standard error of effect size from linear regression. Table S3. Genes that contain or exhibit expression regulation by variants highlighted in GWAS of caudate activity during reward. Germline variants with P GWAS < 0.05 found to be either independent associations by conditional analysis (P conditional < 0.001) or caudate-specific expression quantitative trait loci (eQTL false discovery rate < 0.05) were identified along with their affiliated genes. All eQTL attributes were imported from the Genotype-Tissue Expression project. Table S4. Gene set enrichment analysis results. Gene sets consisting of genes that tend to appear near the top of GWAS results were identified using the correspondences between genes and GWAS P-values established in Table S3. The most significant P-value of a variant contained within each gene or deemed to affect its expression is also listed in parentheses. ES, enrichment score; FDR, false discovery rate; NES, normalized enrichment score. (ZIP 2029 kb)