Summary: Researchers have advanced psychiatric genomics by identifying 641 genes newly associated with schizophrenia. The study combined genetic data from more than 102,000 people with postmortem brain tissue from six cortical regions and employed novel computational models that go beyond traditional local-only mapping. These models capture long-range, network-wide regulatory relationships among genes, revealing coordinated genetic influences that were previously missed.
Rather than limiting analysis to DNA variants adjacent to a gene, the team mapped how distant regulatory variants act across gene co-expression networks—analogous to social networks that link people who live far apart. This network-centered approach demonstrates that long-range genetic coordination across brain regions plays a major role in schizophrenia risk and highlights biological pathways such as glutamate signaling and neuronal development.
Key Facts
- 641 Novel Genes Discovered: By tracing distributed gene communication networks instead of examining genes in isolation, the researchers uncovered hundreds of schizophrenia-linked genes that previous methods overlooked.
- Overcoming the “Lamppost Effect”: Traditional studies typically search only near obvious genetic signals. This work shows most causal regulation occurs at a distance via co-expression networks rather than exclusively through nearby variants.
- Large, Multiregional Dataset: The analysis integrated genotype data from over 102,000 individuals with RNA sequencing from hundreds of donors across six distinct human cortical regions, enabling robust detection of trans-regulatory effects.
- Core Biological Pathways Identified: Newly implicated genes converge on glutamate neurotransmission, synaptic communication, immune-related processes within the brain, and early neurodevelopmental programs—pathways relevant to schizophrenia biology.
- Steps Toward Precision Psychiatry: Moving from single-gene associations to integrated gene-network programs creates a framework for developing targeted interventions tuned to an individual’s specific molecular profile.
Source: Lieber Institute for Brain Development
Background: Schizophrenia is known to have a strong genetic component, but isolating which genes are responsible has been challenging because many disease-associated variants lie outside coding regions and act through complex regulatory networks. Traditional transcriptome-wide association strategies commonly focus on local (cis) regulatory effects, leaving much of gene regulation unexplained.
A multidisciplinary team led by researchers at the Lieber Institute for Brain Development, in collaboration with colleagues from the University of Bari, Italy, and more than 60 psychiatric centers worldwide, developed new analytical tools to address this gap. By integrating genome-wide genotypes with RNA sequencing across multiple brain regions, they modeled how distal (trans) acting variants influence gene expression through co-expression modules.
Published in Nature Genetics, the study introduced two complementary models—INGENE and MODULE—that quantify the combined effect of candidate trans-acting variants within gene co-expression networks. When combined with conventional cis-based predictors, these co-expression-informed models improved gene expression prediction for 18,744 genes across the sampled brain regions and identified 766 genes statistically associated with schizophrenia (P_FDR < 0.01), of which 641 were not previously reported by transcriptome-wide analyses.
“Most genetic studies have been looking under the lamppost—focusing only on variants adjacent to genes,” said Dr. Giulio Pergola, senior author at the Lieber Institute for Brain Development. “By incorporating gene co-expression networks, we expanded the search to capture long-distance regulatory relationships that build the genetic architecture of schizophrenia.”
Dr. Daniel Weinberger, CEO and Director of the Lieber Institute for Brain Development, added: “The work shows schizophrenia risk arises from coordinated gene programs rather than isolated gene effects. Mapping those programs brings us closer to precision psychiatry where interventions can be matched to an individual’s pathway-level biology.”
Key Questions Answered:
A: Earlier approaches often searched only for regulatory variants immediately adjacent to a target gene. Because many regulatory elements act at a distance, those local-focused methods missed trans-regulatory influences—variants in one genomic location that alter the expression of distant genes. By explicitly modeling gene co-expression networks and distal regulation, the new approach uncovers signals outside the traditional search radius.
A: The consortium developed computational models that combine genetic variation with RNA sequencing-derived co-expression networks. Using data from over 102,000 genotyped individuals and RNA-seq from six brain regions, they identified genes that are co-regulated across donors and then assessed how trans-acting variants within those networks predict gene expression. This strategy reveals functional groupings of genes that behave as coordinated units even when they are far apart physically in the genome.
A: The additional genes form clusters that implicate key biological systems rather than a single causal pathway. The most prominent themes include glutamatergic neurotransmission (fast synaptic signaling), synapse formation and communication, local immune and inflammatory processes in the brain, and early neurodevelopmental programs. These pathway-level insights offer practical targets for future mechanistic studies and therapeutic development.
Editorial Notes:
- This article was edited by a Neuroscience News editor.
- The full journal paper was reviewed.
- Additional context was provided by editorial staff.
About this genetics and schizophrenia research news
Author: Gideon Hertz
Source: Lieber Institute for Brain Development
Contact: Gideon Hertz – Lieber Institute for Brain Development
Image: The image is credited to Neuroscience News
Original Research: Open access.
“Co-expression-based models improve eQTL predictions for transcriptome-wide association studies and highlight new schizophrenia-associated genes” by Fabiana Rossi et al., Nature Genetics.
DOI: 10.1038/s41588-026-02646-3
Abstract
Co-expression-based models improve eQTL predictions for transcriptome-wide association studies and highlight new schizophrenia-associated genes
Most genetic variants linked to complex heritable traits lie outside protein-coding regions and are believed to influence disease by regulating gene expression. Standard transcriptome-wide association methods typically emphasize local (cis) genetic effects, leaving many regulatory mechanisms unexplained.
This study demonstrates that including distal (trans) regulatory effects increases the accuracy of gene expression prediction and improves detection of disease-associated genes. Using RNA sequencing from six human postmortem brain regions, the authors developed INGENE and MODULE—two models that capture the combined influence of candidate trans-acting variants within gene co-expression networks.
When integrated with conventional cis-based predictors, these models improved expression imputation (maximum likelihood estimation, α = 0.05) for 18,744 genes across regions. Applied to Psychiatric Genomics Consortium wave 3 genotypes, the framework identified 766 genes associated with schizophrenia (P_FDR < 0.01), including 641 genes not previously reported in transcriptome-wide analyses. These results emphasize the importance of distal regulatory mechanisms and gene network interactions in schizophrenia risk.