Mapping the genetic landscape across 14 psychiatric disorders

Abstract

Psychiatric disorders display high levels of comorbidity and genetic overlap^1,2, challenging current diagnostic boundaries. For disorders for which diagnostic separation has been most debated, such as schizophrenia and bipolar disorder³, genomic methods have revealed that the majority of genetic signal is shared⁴. While over a hundred pleiotropic loci have been identified by recent cross-disorder analyses⁵, the full scope of shared and disorder-specific genetic influences remains poorly defined. Here we addressed this gap by triangulating across a suite of cutting-edge statistical and functional genomic analyses applied to 14 childhood- and adult-onset psychiatric disorders (1,056,201 cases). Using genetic association data from common variants, we identified and characterized five underlying genomic factors that explained the majority of the genetic variance of the individual disorders (around 66% on average) and were associated with 238 pleiotropic loci. The two factors defined by (1) Schizophrenia and bipolar disorders (SB factor); and (2) major depression, PTSD and anxiety (Internalizing factor) showed high levels of polygenic overlap⁶ and local genetic correlation and very few disorder-specific loci. The genetic signal shared across all 14 disorders was enriched for broad biological processes (for example, transcriptional regulation), while more specific pathways were shared at the level of the individual factors. The shared genetic signal across the SB factor was substantially enriched in genes expressed in excitatory neurons, whereas the Internalizing factor was associated with oligodendrocyte biology. These observations may inform a more neurobiologically valid psychiatric nosology and implicate targets for therapeutic development designed to treat commonly occurring comorbid presentations.

Fig. 1. Genome-wide structural models.

a, Heatmap of r_gs across the 14 disorders as estimated using LDSC on the lower diagonal and the correlations among the psychiatric factors as estimated using GenomicSEM above the diagonal. Two-sided P values were derived from the Z-statistics, calculated as the point estimate of the r_g divided by its s.e. Cells depicted with an asterisk reflect values that were significant at a Bonferroni-corrected threshold for multiple comparisons. Exact values are reported in Supplementary Table 1. Disorders that load on the same factor are shown in the same colour. Per the legend at the bottom of the panel, darker blue shading indicates larger, positive r_gs. LDSC estimates were used as the input to genomic SEM to produce the results in b and c. b, Estimates from the five-factor model along with standard errors in parentheses. Estimates are standardized relative to SNP-based heritabilities, where this is equal to the sum of the squared factor loading (the single-headed arrow(s) from the factor to the disorder) and the residual variance (the values on the double-headed arrows on the single-colour circles with text labels that begin with u). Disorders are shown as pie charts; the proportion of residual variance is shaded in purple and the variance explained by the psychiatric factors is shaded in the colour of the corresponding factor. c, Standardized estimates from the p-factor model. The disorders are colour coded as in b, and the first-order factors (F1–F5) are also colour coded to show variance explained by the second-order p-factor in yellow.

Fig. 2. External trait genetic correlations for psychiatric factors.

Point estimates for the r_gs between 14 external traits and the 5 psychiatric factors from the correlated factors model and the p-factor from the hierarchical model. These traits were selected as they were significantly correlated with at least one factor at >0.35 or <−0.35. Bars depicted with a dashed outline were significant for the Q_Trait heterogeneity statistic, which indicates that the pattern of r_gs for that trait did not fit the factor structure. Bars depicted with an asterisk reflect values that were significant at a Bonferroni-corrected threshold for multiple comparisons, that were also not significant at this same Bonferroni corrected threshold for Q_Trait. This is with the exception that the p-factor is depicted with an asterisk even if it is significant for the Q_Trait, provided that the same trait was significantly correlated with the majority (at least three) of the five other factors. The two-sided P values used to evaluate significance were derived from the Z-statistics, calculated as the point estimate of the r_g divided by its s.e. Error bars are ±1.96 s.e., centred around the point estimate of the r_gs. Traits are ordered by the point estimate for the p-factor. The implied sample size for the psychiatric factors was: Compulsive ($\hat{n}$ = 54,100), SB ($\hat{n}$ = 127,202), Neurodevelopmental ($\hat{n}$  = 84,760), Internalizing ($\hat{n}$ = 1,637,337), SUD ($\hat{n}$ = 313,395) and p-factor ($\hat{n}$ = 2,168,621). Sample sizes for the external traits are reported in Supplementary Table 12 and exact P values are reported in Supplementary Table 13.

Fig. 3. Local genetic correlations.

a, An overview of the average patterns of local r_gs across the genome for all pairs of disorders, shown as a heatmap (below diagonal) and a network plot (above diagonal). The colours of the heatmap represent the average local r_gs across all evaluated loci, with darker red and blue shading indicating more negative and positive r_g, respectively; the dot size reflects the strengths of average associations; and the numbers indicate how many of the local r_gs were significant. These results are mirrored in the network plot, where the width or the edges reflect the number of significant associations, meaning that only disorders with at least one significant local r_g are connected, and the edge opacity reflects the strength of the average local r_g across tested loci. Note that label colours are concordant with the genomic SEM factor structure from Fig. 1 and, as shown, disorders of similar colours also tend to be proximally located within the network. b, The local r_g structure within the top r_g hotspot on chromosome (chr.) 11 (112755447–114742317, GRCh37 reference genome), that is, the region where the greatest number of significant r_gs were found across all disorder pairs. Here, the network plot illustrates all significant r_gs detected in this region, with both edge width and opacity reflecting the strength of the association. The region plot in the middle displays the genes contained within the hotspot, and the table below shows the r_g estimates (Rho), 95% confidence intervals (CI^lower, CI^upper), variance explained (R²) and P values for all significant pairwise local r_gs in this region. Label colours are again concordant with those used for the genomic SEM factor structure in Fig. 1.

Fig. 4. Locus-level results.

a, Heatmap of CC-GWAS loci below the diagonal across pairwise combinations of disorders; the darker orange shading indicates a higher number of CC-GWAS hits. CC-GWAS results are not shown for the Internalizing disorders as their r_gs were too high, or for nicotine dependence as this is a continuously measured trait. Genomic SEM results (number of hits and mean χ² for each factor and factor-specific Q_SNP estimate) are reported above the diagonal. Results for the p-factor are shown above the plot along with a Venn diagram of the overlap between p-factor, p-factor Q_SNP and overall CC-GWAS hits. The disorders are ordered and coloured according to the genomic SEM factor structure from Fig. 1. b,c, The Miami and QQ-plots for the p-factor (b) and SBs factors (c), respectively. These panels show the results for the −log₁₀-transformed two-tailed P values for the factor on the top half of the Miami plot and the log₁₀-transformed one-tailed P values for Q_SNP on the bottom half. Factor hits that were within 100 kb of univariate hits are shown as black triangles, novel hits for the factors that were not within 100 kb of a univariate or Q_SNP hit are shown as red triangles and Q_SNP hits are shown as purple diamonds. d, The two-tailed −log₁₀[P] in a Manhattan plot for the CC-GWAS comparison across MD and SCZ, which produced the most hits (orange diamonds), as well as the scatterplot of standardized case–control effect sizes of MD (x axis) versus SCZ (y axis), with CC-GWAS significant SNPs labelled in red. For b–d, the grey dashed lines indicate the significance threshold, which was defined using Bonferroni correction for multiple comparisons.

Fig. 5. Functional annotation of factor variants.

a, GO enrichment analysis of predicted target genes with transdiagnostic associations (that is, variants associated with the p-factor), or those target genes associated with the SB factor that were not overlapping with p-factor target genes. Depicted −log₁₀-transformed P values are one-sided, calculated using a χ² test; false-discovery rate (FDR) correction was applied for multiple comparisons. b, The averaged and normalized expression levels of target genes of the indicated classes along the temporal trajectory of human brain development. Shading around the lines reflects 95% CIs. pcw10, post-conception week 10. c,d, Average log₁₀[P] values across EWCE and MAGMA enrichment for genes associated with the indicated factors in fetal brain cell types using two independent single-cell RNA-sequencing (scRNA-seq) datasets^, (c) or adult brain cell types using three independent single-nucleus RNA-seq (snRNA-seq) datasets^– (d). The P values from EWCE and MAGMA were two-sided and each had an FDR correction applied for multiple comparisons before averaging the two sets of results. EWCE P values were empirically derived using a permutation test; MAGMA P values were calculated using an F-test. Int, Internalizing disorders factor. The implied sample size for the three depicted psychiatric factors was: SB ($\hat{n}$ = 127,202), Internalizing ($\hat{n}$ = 1,637,337) and p-factor ($\hat{n}$ = 2,168,621). CycProg, cycling progenitor; Endo/BBB, endothelial/blood brain barrier; ExNeu, excitatory neuron; InNeu, interneurons; IP, intermediate progenitor; OPC, oligodendrocyte progenitor cell; RG, radial glia; Astro, astrocyte; MSN, medium spiny neuron; ODC/Oligo, oligodendrocyte.

Extended Data Fig. 1. Univariate MiXeR Results.

Power curves estimating the sample size of a GWAS study are needed to saturate the yield of genome-wide significant loci. The legend shows the current effective sample size of today’s GWAS, followed by the projected effective sample size needed for the GWAS yield to saturate.

Extended Data Fig. 2. External trait genetic correlations: Comparison across psychiatric factors.

Bar graphs depict genetic correlations with the 31 complex traits for the five psychiatric factors from the correlated factors model and the second-order, p-factor from the hierarchical model. Panels are separated by the different groupings of traits (e.g., cognitive; socioeconomic). Bars depicted with a dashed outline were significant at a Bonferroni-corrected threshold for the Q_Trait heterogeneity metric that flags traits whose patterns of genetic correlations from LDSC do not conform to those implied by the factor model. Error bars are +/− 1.96 SE that are centred around the point estimate of the genetic correlations. Bar depicted with a * reflect values that were significant at a Bonferroni corrected threshold for multiple comparisons, that were also not significant at this same Bonferroni corrected threshold for Q_Trait. This is with exception of the p-factor, which is depicted with a ‘*’ even if it is significant for the Q_Trait, as long as that same trait was significantly correlated with the majority (at least three) of the five other factors. The two-sided P-values used to evaluate significance were derived from the Z-statistics, calculated as the point estimate of the genetic correlation divided by its standard error. Correlations are ordered according to the point estimate for the p-factor. The implied sample size for the psychiatric factors was: Compulsive ($\hat{n}$ = 54,100); Schizophrenia/Bipolar ($\hat{n}$ = 127,202); Neurodevelopmental ($\hat{n}$  = 84,760); Internalizing ($\hat{n}$ = 1,637,337); Substance Use ($\hat{n}$ = 313,395); p-factor ($\hat{n}$ = 2,168,621). See Suppl. Table 12 for sample sizes for the external traits and Suppl. Table 13 for exact P-values.

Extended Data Fig. 3. External trait genetic correlations: Comparison within factors.

Bar graphs depict genetic correlations with the 31 complex traits that are ordered by magnitude within each factor for the five psychiatric factors from the correlated factors model and the second-order, p-factor from the hierarchical model. Bars depicted with a dashed outline for the Q_Trait heterogeneity metric. Bar depicted with a * reflect values that were significant at a Bonferroni corrected threshold for multiple comparisons, that were also not significant at this same Bonferroni corrected threshold for Q_Trait. This is with exception of the p-factor, which is depicted with a ‘*’ even if it is significant for the Q_Trait, as long as that same trait was significantly correlated with the majority (at least three) of the five other factors. The two-sided P-values used to evaluate significance were derived from the Z-statistics, calculated as the point estimate of the genetic correlation divided by its standard error. Error bars are +/− 1.96 SE that are centred around the point estimate of the genetic correlations. The implied sample size for the psychiatric factors was: Compulsive ($\hat{n}$ = 54,100); Schizophrenia/Bipolar ($\hat{n}$ = 127,202); Neurodevelopmental ($\hat{n}$  = 84,760); Internalizing ($\hat{n}$ = 1,637,337); Substance Use ($\hat{n}$ = 313,395); p-factor ($\hat{n}$ = 2,168,621). See Suppl. Table 12 for sample sizes for the external traits and Suppl. Table 13 for exact P-values.

Extended Data Fig. 4. Stratified Genomic SEM results.

Bar graph depicts the enrichment results for different brain cell types, protein-truncating variant intolerant (PI) genes, and the intersection across PI genes and brain cell types. Results are shown only for the SB, Internalizing, and p-factor due to the limited signal for the other factors. Enrichment for height is depicted in purple to benchmark results and evaluate specificity in signal for the psychiatric factors relative to another human complex trait. Error bars are +/- 1.96 SE that are centred around the enrichment point estimate. Enrichment estimates that were significant at a strict Bonferroni corrected threshold for multiple comparisons are shown with a *. The one-sided P-values used to evaluate significance were derived from the Z-statistics, calculated as the enrichment point estimate divided by its standard error. Exact P-values are reported in Suppl. Table 50. The implied sample size for the psychiatric factors was: Compulsive ($\hat{n}$ = 54,100); Schizophrenia/Bipolar ($\hat{n}$ = 127,202); Neurodevelopmental ($\hat{n}$  = 84,760); Internalizing ($\hat{n}$ = 1,637,337); Substance Use ($\hat{n}$ = 313,395); p-factor ($\hat{n}$ = 2,168,621).