📄 genetic and genomic medicine

Identifying and Characterising Common Genetic Differences in Schizophrenia and Bipolar Disorder

By applying Case-Case GWAS to over 100,000 cases, this study identified 19 genome-wide significant loci with divergent genetic effects that differentiate schizophrenia from bipolar disorder, revealing distinct neurodevelopmental pathways and biological mechanisms despite their substantial shared genetic liability.

Original authors: Willcocks, I. R., Richards, A., Legge, S. E., Holmans, P., Di Florio, A., Cardno, A. G., O'donovan, M. C., Owen, M. J., Pardinas, A. F., Walters, J. T.

Published 2026-07-19

📖 6 min read🧠 Deep dive

CC BY 4.0

Original authors: Willcocks, I. R., Richards, A., Legge, S. E., Holmans, P., Di Florio, A., Cardno, A. G., O'donovan, M. C., Owen, M. J., Pardinas, A. F., Walters, J. T.

Original paper licensed under CC BY 4.0 (https://creativecommons.org/licenses/by/4.0/). ⚕️ This is an AI-generated explanation of a preprint that has not been peer-reviewed. It is not medical advice. Do not make health decisions based on this content. Read full disclaimer

Imagine your brain is like a massive, bustling city. For decades, scientists have been trying to map out why some people get stuck in traffic jams called "Schizophrenia" while others get stuck in different kinds of traffic called "Bipolar Disorder." These two conditions are like neighboring cities that look very similar from a distance; they share the same skyline, the same weather patterns, and even many of the same buildings. In fact, they share so much genetic "real estate" that for a long time, researchers thought they were just different versions of the same problem.

But here's the twist: even though these cities share a lot of DNA, they have distinct neighborhoods that make them unique. To find these differences, scientists use a tool called a Genome-Wide Association Study (GWAS). Think of this as a giant spotlight scanning millions of tiny switches (genes) in your DNA to see which ones are flipped "on" in sick people compared to healthy people. Usually, they compare sick people to healthy people. But to find what makes Schizophrenia different from Bipolar Disorder, you need a special trick: you have to compare the sick people of City A directly against the sick people of City B. This is like asking, "If you live in the Schizophrenia traffic jam, what specific road signs do you see that you don't see in the Bipolar traffic jam?" This study does exactly that, looking for the genetic "road signs" that point to one city but not the other.

The Big Hunt: Finding the Genetic Differences

In this study, a team of researchers acted like genetic detectives. They gathered a massive crowd of 67,390 people with Schizophrenia and 41,917 people with Bipolar Disorder. Instead of comparing them to healthy people, they compared the two groups directly to each other. They were looking for the genetic "fingerprint" that separates these two conditions.

What they found was a treasure trove of 19 specific locations in the human genome that act as differentiators. Imagine these 19 spots as unique street corners. At 16 of these corners (84% of them), the genetic signal was completely opposite. It's as if a specific genetic switch acts like a "Go" sign for Schizophrenia but a "Stop" sign for Bipolar Disorder. This is a huge deal because it proves that these two disorders aren't just different degrees of the same thing; they have genuine, opposing genetic forces at play.

The researchers also found 4 of these special corners that were so subtle, they were invisible when looking at Schizophrenia or Bipolar Disorder alone. You needed to look at them side-by-side to see the difference. This shows that the "compare-the-sick" method is a super-powerful magnifying glass that can spot clues that other methods miss.

What Do These Genes Do?

Once they found these 19 spots, the team asked: "What kind of biological machinery lives here?" They mapped these locations to 102 genes. The results were very clear: these genes are almost exclusively active in the brain. They didn't show up in the liver, the heart, or the skin. It's like finding a secret club that only meets in the library and never in the gym.

When they looked closer at what these genes actually do, they found a strong theme: neurons and synapses. Think of neurons as the wires in your brain and synapses as the connections where those wires talk to each other. The study suggests that the genes making Schizophrenia unique are heavily involved in how these wires are built and how they connect. It's like finding that the blueprints for Schizophrenia have a lot of extra notes about how to lay the foundation and wire the house, whereas Bipolar Disorder's blueprints might be more focused on the furniture inside. This supports the idea that Schizophrenia might be more deeply rooted in how the brain develops early in life.

The "Fingerprint" in Real Life

To see if these genetic differences actually matter in the real world, the researchers tested them in a group of 817 people with Schizophrenia. They built a "genetic score" based on those 19 special spots.

The score told a clear story:

Timing: People with a higher score for these specific Schizophrenia genes tended to get sick earlier in life. It's like having a genetic head start on the illness.
Symptoms: The score was linked to more severe "negative symptoms" (like feeling flat, having less energy, or not showing emotion) and "disorganized symptoms" (like confused thinking). Interestingly, the genetic score for Bipolar Disorder didn't predict these specific problems. This suggests that the unique genetic part of Schizophrenia is what drives these specific, tough-to-treat symptoms.

What They Ruled Out

The researchers were careful to make sure they weren't just finding things that both disorders share. They used a special filter to remove any genetic spots that seemed to cause risk for both conditions in the same way. They found that while the two disorders share a lot of genetic ground (about 68% of their risk is shared), the things that make them different are mostly these "opposite" signals, not just stronger or weaker versions of the same signal.

They also noted that their findings were mostly about Schizophrenia. Because the Schizophrenia group in their study was larger and had more data, they found more differences that pointed toward Schizophrenia than Bipolar Disorder. They admit that if they had a bigger group of Bipolar patients, they might find more unique "Bipolar-only" spots.

The Bottom Line

This study confirms that while Schizophrenia and Bipolar Disorder are close cousins in the genetic family, they have distinct personalities. The differences aren't just a matter of "how much" risk you have, but "which way" the risk points. The unique genetic factors for Schizophrenia seem to be deeply tied to how the brain's wiring is built and how it connects, potentially explaining why the illness starts earlier and includes those specific, hard-to-treat symptoms. It's a step forward in understanding that these are two different biological stories, even if they sometimes look like they're telling the same plot.

Technical Summary: Identifying and Characterising Common Genetic Differences in Schizophrenia and Bipolar Disorder

Problem Statement
Schizophrenia (SCZ) and Bipolar Disorder (BD) are distinct diagnostic categories that share substantial clinical features and a high degree of genetic overlap (genetic correlation $r_g \approx 0.68$ ). While previous research has extensively mapped shared genetic liability, there is a critical need to identify specific genetic variants that differentiate these disorders. Understanding these differences is essential for elucidating the biological processes that drive distinct clinical presentations, potentially informing biomarker identification, clinical subtyping, and the development of novel therapeutics. Previous attempts to identify these differentiating alleles have been limited by small sample sizes, reliance on individual-level data (which is often restricted), or a focus on shared rather than divergent effects.

Methodology
The authors employed a Case-Case Genome-Wide Association Study (CC-GWAS) approach to identify common genetic variants differentially associated with SCZ and BD.

Data Sources: The analysis utilized summary statistics from the largest available GWAS for SCZ (67,390 cases, 94,015 controls; PGC22) and BD (41,917 cases, 371,549 controls; PGC24). The authors prioritized phenotypic specificity over sample size for BD, selecting the 2021 PGC GWAS over a newer, larger study based on self-reported diagnoses.
CC-GWAS Implementation: Using the CC-GWAS method, the study compared allele frequency differences between the two case groups, adjusting for control overlap (45,497 shared controls), disorder prevalence, SNP-based heritability, and genetic correlation. Positive betas indicated alleles conferring greater risk for SCZ relative to BD.
Locus Refinement: Initial genome-wide significant (GWS) loci were filtered to isolate "disorder-differentiating" signals. Loci showing evidence of shared liability (convergent effects) were excluded using Genomic Structural Equation Modeling (GenomicSEM) to ensure the final set represented true divergence. Two MHC region loci were also removed due to known shared risk.
Downstream Analyses:
- Functional Annotation: FUMA was used for SNP-to-gene mapping (positional and eQTL) across 13 brain regions and PsychENCODE.
- Enrichment: MAGMA was used for gene set enrichment analysis (GSA) and tissue expression enrichment.
- Heritability & Correlation: LD Score Regression (LDSC) estimated disorder-differentiating heritability and genetic correlations with other psychiatric and cognitive traits.
- Polygenic Risk Scores (PRS): PRS derived from the CC-GWAS results were tested in an independent SCZ cohort (CardiffCOGS) for associations with clinical phenotypes (age at onset, symptom severity, treatment resistance).

Key Contributions and Results

Discovery of Divergent Loci: The study identified 19 genome-wide significant loci differentiating SCZ and BD. Crucially, 16 of these (84%) demonstrated divergent genetic effects, where risk alleles showed opposite directions of association between the two disorders. Only three loci showed convergent effects (same direction, larger effect in SCZ).
Enhanced Power of CC-GWAS: Four of the 19 loci reached genome-wide significance in the CC-GWAS despite not attaining significance in either individual disorder GWAS. This demonstrates the method's enhanced power to detect alleles with opposite-direction effects that are missed by standard case-control analyses.
Biological Characterization:
- Tissue Specificity: The 102 genes mapped to these loci showed significant expression enrichment across all 13 tested brain regions (including frontal cortex and anterior cingulate) but no significant enrichment in peripheral tissues.
- Pathways: Gene set enrichment analysis identified "Neuronal Projection" and "Synapse" as the strongest biological themes differentiating the disorders, supporting neurodevelopmental hypotheses.
Genetic Correlations: The disorder-differentiating signal showed a strong positive genetic correlation with SCZ ( $r_g = 0.79$ ) but a near-zero correlation with BD ( $r_g = 0.098$ ). Notably, while both SCZ and BD generally show positive correlations with educational attainment, the CC-GWAS signal showed a negative correlation, indicating that SCZ-specific risk alleles are associated with lower educational attainment and cognitive performance.
Clinical Associations: In the CardiffCOGS cohort, the CC-GWAS PRS was significantly associated with:
- Earlier age at onset of psychosis.
- More severe negative symptoms (diminished expressivity).
- More severe disorganised symptoms.
  These associations mirrored the SCZ PRS but were absent for the BD PRS, suggesting SCZ-specific variants mark neurodevelopmental dimensions of illness.

Significance
The paper claims to provide concrete targets for understanding the pathogenic differences between SCZ and BD, moving beyond the substantial shared liability that characterizes these conditions. The findings demonstrate that genuine divergent genetic effects exist, predominantly characterized by opposite-direction associations rather than just quantitative differences in effect size. The convergence of these divergent loci on neuronal and synaptic processes, combined with their association with cognitive impairment and earlier onset, supports the hypothesis that SCZ-specific genetic risk is characterized by greater neurodevelopmental disruption compared to BD. The study validates CC-GWAS as a powerful tool for leveraging publicly available summary statistics to uncover disorder-specific biology that might otherwise remain undetected.

More like this