Neurobehavioural genes exert indirect genetic effects on blood lipid profiles through behavioural traits

Figures

• 

• 

  Fig. 1. A flowchart of analyses undertaken in this project. A family-based study was conducted to explore the IGEs and DGEs on lipid profiles. (1) Identify genetic variants with IGEs and DGEs. (2) Compare genomic distribution, effect directions, and effect sizes between IGE- and DGE-variants. (3) Identify prioritized genes of IGEs and DGEs based on V2G, PoPS, and TWAS. (4) Functional annotation, tissue enrichment, and pathway enrichment. (5) Explore the potential behavioral factors contributing to lGEs. FISSIC Fangshan Family-based Ischemic Stroke Study in China, IGEs indirect genetic effects, DGEs direct genetic effects, V2G variant-to-gene, PoPS gene-level polygenic priority scores, TWAS multitissue transcriptome-wide association study.

• 

  Fig. 2. Multi-layered Manhattan plot of within-sibship genome-wide association for IGEs, DGEs, and TGEs. A Multi-layered Manhattan plot for HDL-C. B Multi-layered Manhattan plot for LDL-C. C Multi-layered Manhattan plot for TC. D Multi-layered Manhattan plot for TG. The red, blue, and green rings represent variants exhibiting DGEs, IGEs, and TGEs, respectively, arranged from the outermost to the innermost layer. The x-axis represents chromosomal position. The y-axis represents –log₁₀(P) values of DGEs, IGEs, and TGEs from the within-sibship genome-wide association. Yellow dots indicate variants exceeding the significance threshold of P < 5 × 10⁻⁶, while red dots indicate variants reaching genome-wide significance (P < 5 × 10⁻⁸). Chr chromosome, HDL-C high-density lipoprotein cholesterol, LDL-C low-density lipoprotein cholesterol, TC total cholesterol, TG triglycerides.

• 

  Fig. 3. Genetic loci with IGEs, DGEs, and TGEs for lipid profiles. A Map of genetic loci exhibiting IGEs, DGEs, and TGEs on chromosomes for lipid profile. Colored dots represent different effect types: blue for IGEs, red for DGEs, and green for TGEs. Labels adjacent to the points denote chromosomal bands. The symbols next to the chromosomal band names indicate validation status: * indicates the locus passed trio-stage validation, # indicates DGEs did not pass, while TGEs passed validation. B The number of genetic loci exhibiting IGEs, DGEs, and TGEs per lipid phenotype category. The bar plot in the bottom left shows the number of loci exhibiting IGEs (blue), DGEs (red), and TGEs (green) across lipid traits. In the top bar plot, gray bars indicate loci shared by two traits, while bars without intersections represent trait-specific loci. HDL-C high-density lipoprotein cholesterol, LDL-C low-density lipoprotein cholesterol, TC total cholesterol, TG triglycerides, IGEs indirect genetic effects, DGEs direct genetic effects, TGEs total genetic effects.

• 

  Fig. 4. Comparison of effect directions and effect sizes of IGEs and DGEs. A Different genetic effects (IGEs, DGEs, and TGEs) of the 20 leading SNPs exhibiting IGEs. The x-axis shows the estimated effect sizes (β) for each type of genetic effect, as determined by the within-sibship genome-wide association analysis (n = 771). The colors of the dots indicate the types of effects: blue for IGEs, red for DGEs and green for TGEs. Dot size corresponds to statistical significance, with larger dots indicating stronger significance. From largest to smallest, dot sizes reflect the following P value thresholds: P < 5 × 10⁻⁸, P < 0.05 and P > 0.05. B Comparison of effect directions for IGEs, DGEs, and TGEs. It summarizes the proportion of SNPs showing consistent directions of effect between pairs of genetic effect types: IGEs, DGEs, and TGEs. The y-axis represents the proportion. IGEs vs. DGEs: The proportion of SNPs for which IGEs and DGEs exhibit consistent directions of effect. IGEs vs. TGEs: The proportion of SNPs for which IGEs and TGEs exhibit consistent directions of effect. DGEs vs. TGEs: The proportion of SNPs for which DGEs and TGEs exhibit consistent directions of effect. C Comparison of…

• 

  Fig. 5. Prioritized genes with different genetic effects (IGEs, DGEs, and TGEs). A Schematic diagram of gene prioritization. B Circular heatmap showing predictor score for candidate genes. Prioritized genes within each locus are highlighted in red. Grid intensity represents a weighted composite score of V2G, PoPS, and TWAS statistics (weight ratio 2:2:1). C The number of prioritized genes. The bar plot in the lower-left corner shows the number of genes exhibiting IGEs (blue), DGEs (red), and TGEs (green). The bar plot on the upside depicts the number of genes with different genetic effects. Gray bars indicate genes exhibiting both DGEs and TGEs, while green, red, and blue bars represent genes uniquely exhibiting TGEs, DGEs, and IGEs, respectively. The y-axis shows the number of genes. IGEs indirect genetic effects, DGEs direct genetic effects, TGEs total genetic effects, V2G variant-to-gene, PoPS gene-level polygenic priority scores, TWAS multitissue transcriptome-wide association study.

• 

  Fig. 6. Tissue enrichment and pathway analysis of prioritized genes exhibiting IGEs and DGEs. A Tissue enrichment of prioritized genes exhibiting IGEs. Each horizontal bar represents a different tissue type, with the length of the bar corresponding to the –log₁₀ P-value. A total of 54 tissues from the GTEx v8 were included in this analysis. Source data available as Supplementary Data 32. B Pathway enrichment of genetic variants exhibiting IGEs estimated by within-sibship genome-wide association analysis. The length of the bars represents the adjusted –log₁₀(P adjust) values, while the position of the dots indicates the corresponding Z-statistic. The pathway categories were derived from the Gene Ontology (GO) resource and the Kyoto Encyclopedia of Genes and Genomes (KEGG), and include biological process (BP), cellular component (CC), and molecular function (MF). Pathways of BP, CC, and MF were represented by pink, yellow, and green, respectively, while KEGG pathways were shown in blue. The number of dots on the left indicates the number of genes enriched in each pathway. From largest to smallest, the dot sizes correspond to gene counts of 200, 150, 100, and 50, respectively. Source data…

• 

  Fig. 7. The potential effects of behavioral factors contributing to lGEs. A The associations between 20 leading IGE-SNPs and lipid profiles before and after adjustment with behavioral factors in the FISSIC cohort (n = 771). Points and lines represent the estimated IGEs with 95% confidence intervals. Source data available as Supplementary Data 36. B The associations between IGE-GRS and lipid profiles before and after adjustment with behavioral factors in the FISSIC cohort. Points and lines represent the estimated IGEs with 95% confidence intervals (n = 1661). C The associations between IGE-GRS and lipid profiles before and after adjustment with behavioral factors in UK Biobank participants (n = 499,849). Points and lines represent the estimated IGEs with 95% confidence intervals. D The mediation analysis between IGE-GRS and lipid profiles (n = 1661). Points and lines represent the estimated average causal mediation effects (ACME) and corresponding 95% confidence intervals for each mediating factor included in the mediation analyses. E The mediation analysis between DGE-GRS and lipid profiles (n = 1661). Points and lines represent the estimated ACME and corresponding 95% confidence…