Schizophrenia, bipolar disorder, and major depressive disorder show overlapping features and differ in age at onset, cognitive impact, impairment severity, and clinical course. Because psychosis can occur across these diagnoses, the authors model psychosis as a transdiagnostic spectrum and use genomic data to probe functionally distinct components within that spectrum. The goal was to capture both shared and specific genetic liabilities that map onto clinically meaningful symptom domains.
The study applied Genomic Structural Equation Modeling (Genomic SEM) to summary-level genome-wide association results for schizophrenia, bipolar disorder, and major depression. From those data the authors derived four latent factors intended to index distinct psychosis-related dimensions shared across the three disorders:
These factors represent a clinically informed decomposition of genetic liability across disorders, allowing interrogation of biological annotations and developmental trajectories for each component.
To understand biological specificity, the investigators used Stratified Genomic SEM to test enrichment across functional annotations and brain cell types and regions. In parallel, they conducted transcriptomic analyses to assess developmental timing and expression trends associated with each latent factor. The analyses leveraged publicly available GWAS summary statistics and developmental transcriptomic resources cited in the study’s methods.
The factors exhibited differing genetic correlations with personality dimensions. Notable findings reported were:
These contrasts suggest that distinct psychosis-spectrum components relate to personality traits in different directions, consistent with clinical heterogeneity.
Across the psychosis continuum, several enrichments were significant:
The study also reported factor-specific enrichment patterns:
These results indicate overlap in implicated neural substrates across factors while highlighting unique cellular and regional signals tied to particular genetic components.
Transcriptomic profiling across developmental stages revealed different temporal expression trends by factor:
Together, these timing differences support a partial neurodevelopmental origin for some psychosis-spectrum components while other components show more temporally extended expression.
The authors conclude that a clinically informed genomic factor model can clarify the functional heterogeneity of the psychosis spectrum by separating shared from specific biological signals. The approach identifies both overlapping and factor-specific cellular, regional, and developmental signatures, which could inform future mechanistic studies and refine biological hypotheses about symptom domains across disorders.
Limitations noted in the article context include that the work is based on summary genetic data and transcriptomic resources and that the report is a preprint that has not been peer reviewed. Specific numerical results, sample sizes, p values, and full methodology details are provided in the full manuscript and supplementary materials; those exact values were not reproduced in this summary.
The study relied on publicly available GWAS and transcriptomic datasets (links and sources are listed in the author declarations). The authors state that all data produced are available upon reasonable request. They declared no competing interests and confirmed ethical compliance consistent with use of public human data. The preprint explicitly notes it has not been certified by peer review and should not guide clinical practice.