Psychiatric Genomics · Polygenic Architecture

Shared Genetic Foundations of
Schizophrenia & Bipolar Disorder

Large-scale GWAS reveal a deeply overlapping polygenic landscape — thousands of common variants each exerting small additive effects across both conditions.

~270Genome-wide significant loci in schizophrenia (PGC3, 2022)
~64Genome-wide significant loci in bipolar disorder (PGC BPD3, 2021)
rg ≈ 0.70Genetic correlation between SCZ and BPD
>100KCommon SNPs contribute meaningfully to polygenic risk scores
Genetic Overlap Architecture
SCZ Schizophrenia BPD Bipolar Disorder ~200 unique loci ~70+ shared loci ~30 unique loci h² ≈ 80% (SNP h²: 23%) h² ≈ 75% (SNP h²: 19%)
Genetic Correlation & Representative Effect Sizes
0.70 genetic correlation 0 1

LD-score regression rg
SCZ ↔ BPD

Top locus odds ratios
CACNA1COR 1.15
ANK3OR 1.18
MHC / HLA regionOR 1.22
NRXN1OR 1.12
GRIN2AOR 1.10

Most common variants have small individual effects (OR 1.05–1.25). Polygenicity arises from their large number acting additively.

Schematic Manhattan Profile — Key Chromosomal Regions
● SCZ-specific ● BPD-specific ● Shared — — GWS threshold (p < 5×10⁻⁸)
Heritability Decomposition
Schizophrenia h²=80% SNP h²≈23% Bipolar Disorder h²=75% SNP h²≈19% Common SNP h² Other variance

Total heritability exceeds common-SNP h², consistent with rare variants, CNVs, and gene-environment interaction also contributing.

Notable Shared & Disorder-Specific Loci
Gene / RegionChr.AssociationBiological Role
CACNA1C12p13BothL-type Ca²⁺ channel; neuronal excitability
ANK310q21BothAnkyrin-G; axon initial segment scaffolding
MHC / HLA6p21SCZComplement C4A; synaptic pruning excess
NRXN12p16BothNeurexin-1; synaptic adhesion & organization
GRIN2A16p13SCZNMDA receptor subunit GluN2A; glutamate
SYNE16q25BPDNesprin-1; nuclear envelope & cerebellar signaling
DISC11q42BothDisrupted in schizophrenia 1; neurodevelopmental hub
DRD211q23SCZDopamine D₂ receptor; primary antipsychotic target
Enriched Biological Pathways
Glutamate signalingCalcium channel activitySynaptic transmissionNeuronal development Dopaminergic synapseComplement cascadeSynaptic pruningNMDA receptor complex Circadian regulationHCN channel activitycAMP / PKA pathwayAxon guidance
Cross-disorder PRS prediction
SCZ PRS → BPD risk elevationr² ≈ 0.28
BPD PRS → SCZ risk elevationr² ≈ 0.24
GWAS Discovery Timeline — Schizophrenia & Bipolar Disorder
2009
First reproducible SCZ GWAS loci identified; MHC region implicated by Psychiatric GWAS Consortium (PGC).
2011
BPD GWAS confirms CACNA1C and ANK3 as robust, independently replicated susceptibility loci.
2013
Cross-Disorder Group (CDG) establishes substantial shared genetic architecture across 5 psychiatric conditions including SCZ & BPD (rg ≈ 0.60–0.80).
2014
PGC SCZ mega-analysis: 108 GWS loci. DRD2 and complement C4A implicate dopamine and synaptic pruning.
2019
PGC BPD2 (32k cases): 30 GWS loci identified; genetic differentiation of BPD-I from BPD-II first demonstrated.
2021–22
PGC BPD3 (41k) & PGC3 SCZ (76k): 64 and ~270 GWS loci. rg ≈ 0.70 robustly confirmed. Rare CNVs add independent risk layers.
Present
Multivariate genomic SEM, fine-mapping, and functional annotation refine shared causal variants. PRS explain ~4–9% of liability variance in held-out samples.
Horizon
Ancestrally diverse biobanks and single-cell genomics aim to resolve cell-type-specific mechanisms and improve global transferability of polygenic scores.

Sources: PGC SCZ3 (Trubetskoy et al., Nature 2022) · PGC BPD3 (Mullins et al., Nature Genetics 2021) · Cross-Disorder Group (Anttila et al., Science 2018; Lee et al., Nature Genetics 2019)
h² = twin-based heritability · SNP h² = common-variant narrow-sense heritability · rg = genetic correlation · OR = odds ratio · GWS = genome-wide significant (p < 5×10⁻⁸) · CNV = copy number variant