Catenin Alpha 2 May Be a Biomarker or Potential Drug Target in Psychiatric Disorders with Perseverative Negative Thinking

Figures

• 

  Figure 1. Manhattan plots for the associations of each CTNNA2 SNP with each outcome variable. (A–E) display cardiovascular risk phenotypes, (F–O) display BSI psychiatric symptom scores, and (P–R) display RRS rumination, brooding and reflection scores, respectively. The p-value is displayed with two thresholds, Bonferroni-corrected gene-wide and nominal significance, as a function of the SNP position within the gene. We can see that, while no SNP survived the gene-wide significant threshold, there are nominally significant associations for all 18 phenotypes. R2—linkage disequilibrium with the most significant SNP (marked with asterisk and given by name); SNP—single-nucleotide polymorphism; chr—chromosome; BMI—body mass index; CVD—cardiovascular disease; CHD—coronary heart disease; HCHD—hard coronary heart disease: myocardial infarction or coronary death; BSI—Brief Symptom Inventory; RRS—Ruminative Response Scale.

• 

  Figure 2. The resulting network model from the Bayesian Multilevel Analysis. The edges (all lines) represent a probabilistic relationship between the variables, namely single variables and categories of variables. Single variables include the 18 phenotypes, sex, age, and the two SNPs. Categories of variables include Framingham, RRS, and BSI. The thickness of the edges is proportional with the posterior probability, signifying a more probable real-life connection between the variables (numerical posterior probability values are detailed in Supplementary Table S3). The black edges connect single variables with each other, and the colored edges connect single variables with the categories, where the color encoding is as follows: red—Framingham; yellow—RRS; blue—BSI. The length of the horizontal bars below the nodes represents the posterior probability values with respect to the categories. The results suggest a real-life relationship between rs17019243 and RRS, and BSI single variables and categories as well. However, rs12615043 has no real-life relationship with any of them. RRS—Ruminative Response Scale; BSI—Brief Symptom Inventory; BMI—body mass index; CVD—cardiovascular disease;…

• 

  Figure 3. Structural equation modeling to test the direct and indirect effects of CTNNA2 rs17019243 for BSI psychiatric symptom scores via RRS rumination within the same model. Model fit indices: χ2 = 20.184 (df = 14; p = 0.1244); RMSEA = 0.026 (PCLOSE = 0.949); CFI = 0.999; TLI = 0.992. Direct effects of rs17019243 are displayed in the figure, and the significant (p ≤ 0.05) one is marked in bold. The results suggest that rs17019243 has a direct effect on RRS rumination, but has no direct effect on any of the BSI scores. BSI—Brief Symptom Inventory; RRS—Ruminative Response Scale; df—degrees of freedom; RMSEA—root mean squared error of approximation; PCLOSE—significance of the statistical hypothesis that RMSEA is different from the desirable ≤0.05; CFI—comparative fit index (>0.90 indicates a good model fit, and >0.95 indicates a very good model fit); TLI—Tucker−Lewis index (>0.90 indicates a good, and >0.95 a very good model fit).