Research study · Brain and Behavior · 2026
Multi‐Trait Genetic Insights Into Schizophrenia Across Ancestries: Genome‐Wide Association Meta‐Analyses, Machine Learning, and Drug Repurposing Study
Verified on PubMed Systematic review or meta-analysis
A 2026 meta-analysis in Brain and Behavior in schizophrenia. The results point to shared genetics, but they do not show that any medicine helps or harms people with schizophrenia. This study shows that genetic patterns linked to schizophrenia overlap with patterns seen in people who take certain common medicines, but it does not show that.
In plain language
Study overview
Researchers used large collections of genetic data from people with schizophrenia, a serious mental illness, and from people in several populations, and compared them with genetic data about the medicines people take. They then used computer methods to look for shared genetic patterns and possible drug targets. The study used existing records and computer analysis and did not treat or follow any patients.
Findings
The results point to shared genetics, but they do not show that any medicine helps or harms people with schizophrenia. The study found that some genetic patterns linked to schizophrenia also appear with the use of certain heart, blood pressure, diabetes, and cholesterol medicines. Its computer tools also suggested a few groups of drugs, including ones that act on the body's cannabis-like signaling system, as ideas worth testing in real studies.
The deeper readScientific analysis
Background: Schizophrenia is a psychiatric disorder characterized by a wide range of symptoms and a complex interplay between genetic and environmental factors. Despite advances in genomics, the genetic underpinnings of SC and its relationship with drug use remain poorly understood. Methods: This study utilized genome-wide association data from 53,386 SC patients and 77,258 controls of mainly European ancestry, and 22,778 patients and 35,362 controls of mainly East Asian ancestry, alongside GWAS summary data for 23 types of drug use. Genetic correlation analysis was conducted using linkage disequilibrium score regression, followed by meta-analysis of genome-wide association studies across multiple traits related to SC. Independent risk loci were identified, and their biological significance was explored using molecular methodologies. Machine learning techniques were employed to develop predictive models for SC diagnosis, and drug repurposing opportunities were investigated. Results: We found significant genetic correlations between SC and drug use in both ancestries, uncovering 24 previously unidentified loci associated with SC. Through various analytical strategies, we highlighted novel genes linked to SC and developed diagnostic models. Drug repurposing analysis identified ten drug classes with potential therapeutic relevance for SC, including cannabinoid receptor antagonists, which have shown promise in alleviating symptoms. Conclusion: Our findings reveal a common genetic foundation linking susceptibility to schizophrenia with drug use, providing new insights into the genetic architecture of SC across ancestries. The identification of novel loci and genes associated with SC enhances our understanding of its biological underpinnings and offers potential targets for therapeutic intervention. Additionally, our study opens avenues for drug repurposing as a promising strategy for SC treatment, underscoring the importance of integrating genetic and pharmacological research in psychiatry.
The paper
Citation
Yan P, Wang H, Shu Y, Xie R, Reyidan A, Fu S Multi‐Trait Genetic Insights Into Schizophrenia Across Ancestries: Genome‐Wide Association Meta‐Analyses, Machine Learning, and Drug Repurposing Study. Brain and Behavior (2026). PubMed 42421611 · DOI
Limitations the authors noted
Relied on summary-level GWAS data, limiting modelling of individual-level interactions among genetics, medication exposure, and environment. Drug-use GWAS traits reflect a mix of medication exposure, underlying disease liability, healthcare access, and prescribing behavior. Cross-ancestry and multi-trait methods may introduce heterogeneity from phenotype definition, LD structure and sample composition. Gene prioritization, TWAS, SMR, proteomic, and CMap analyses are informative but do not establish causality; experimental validation is required. Machine learning and transcriptomic validation used small, pooled datasets and results are described as supportive rather than diagnostic.
In practice
Why this matters
This study shows that genetic patterns linked to schizophrenia overlap with patterns seen in people who take certain common medicines, but it does not show that those medicines cause or prevent the illness. Its suggestion that cannabinoid-related drugs might be worth studying is a computer-generated idea, not evidence that any product helps. Whether these findings lead anywhere depends on future laboratory and clinical trials in other populations, with direct measurement of symptoms over time.
Browse related research
Condition: Genetic predisposition to disease Condition: Linkage disequilibrium Condition: Schizophrenia
Related on CED Clinic
Back to Daily Research and Patient Questions
Reviewed by Dr. Caplan on October 14, 2026.
This page is general education about a published study. It is not medical advice and does not create a doctor-patient relationship. For care specific to you, book a consultation.
