Positive selection on brain cis-regulatory elements in the human lineage drives changes in gene expression and susceptibility to neuropsychiatric disorders
Youness Touissi, Eric J Vallender · 2026
Read Positive selection on brain cis-regulatory elements in the human lineage drives changes in gene expression and susceptibility to neuropsychiatric disorders by Youness Touissi, Eric J Vallender (2026) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/pmc-13266179.txt.
Introduction
Youness Touissi and Eric Vallender scanned human brain gene switches in two thousand twenty-six. Human brain change likely ran through shifts in gene control. Such changes alter how genes work, not proteins themselves. The team used hundreds of primate genomes. They tested brain-active switches for signs of selection found only in humans. Thousands of switches showed positive selection signals.
Genes near those switches were not the same in activity versus macaques. Gaps showed in fetal and adult stages. Always-on switches held more psychiatric risk variants. Schizophrenia and bipolar disorder links stood out. Evolution may have traded brain power for weakness. The work connects those roots to modern mental illness.
Catalog record
| CREATOR | Youness Touissi, Eric J Vallender |
|---|---|
| YEAR | 2026 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 47 KB |
| VERSION | 1.0 |
| LICENSE | CC BY 4.0 |
| RIGHTS | CC BY |
| SOURCE | Europe PMC |
Browse the full library · About the Tower of Babel
Full text
The opening of the work, from the library's hosted text:
Positive selection on brain cis-regulatory elements in the human lineage drives changes in gene expression and susceptibility to neuropsychiatric disorders Authors: Youness Touissi, Eric J Vallender Published: 2026 ABSTRACT SummaryHuman brain evolution is hypothesized to be driven primarily by regulatory rather than protein-coding changes. Identifying evolutionary changes in regulatory elements, however, has been constrained by numbers and relationships of taxa and by methodology targeting conserved sequences. These approaches are independent of any experimental assessment of function and overlook potential neofunctionalization in less constrained regions. Here, we leveraged 243 primate genomes to apply a human branch-specific positive selection test to over 350,000 experimentally confirmed human brain-active cis-regulatory elements, identifying 8,522 elements under positive selection on the human lineage. Genes regulated by human positively selected elements showed differential expression between humans and macaques during fetal and adult stages compared to non-selected elements. Elements under positive selection that remained active across fetal and adult stages showed the strongest enrichment for genetic variants associated with neuropsychiatric conditions, including schizophrenia and bipolar disorder.Keywords: brain evolution, comparative genomics, cis-regulatory elements, HARs, positive selection, psychiatric disorders Introduction Human brain evolution has involved structural and functional changes resulting in unique cognitive abilities compared to other primate species.1 Interestingly, this specialization is largely not correlated with major differences in protein-coding genes, particularly compared to our closest ape relatives.2 Instead, changes in how genes are regulated appear to drive the key differences in human brain development.2 This evolution of gene regulation is driven by changes in the non-coding parts of the genome3 containing elements, such as enhancers and promoters, that orchestrate spatial and temporal expression during development.4,5 The modular nature of cis-regulatory elements (CREs) drives evolutionary innovation by allowing gene expression changes in tissues, cell types, or developmental time points without producing deleterious pleiotropic effects elsewhere.6 This regulatory evolution is common across various contexts but is most evident in the brain, which shows significant enrichment for human-specific CRE activity compared to other tissues.7,8,9 The evolution of the human brain has been associated with the unique susceptibility of humans to a range of neuropsychiatric disorders, particularly those with neurodevelopmental origins.7,10 Genome-wide association studies (GWASs) consistently show that more than 90% of genetic variation linked to disorders such as schizophrenia and autism are located in the non-coding, regulatory parts of the genome.11,12 Human-evolved regulatory elements, such as human gained enhancers (regulatory regions that gained activity in humans) and human accelerated regions (conserved regions that accumulated sequence changes in the human branch), are significantly enriched near genes associated with autism or schizophrenia risk.7 Genetic variation in regulatory elements active during brain development account for a substantial portion of inherited differences in cognitive and psychiatric traits.13 Together, these convergent lines of evidence suggest that studying the evolution of brain-relevant CREs is essential for a better understanding of the genetic basis of both human brain evolution and psychiatric disorders.14 Previous efforts to find functionally significant non-coding regions have focused on finding highly conserved sequences across species, inferring functional importance,15,16,17 and identifying regions where the human lineage is more rapidly diverging as evidence of positive selection pressure.8,18,19 Large-scale comparative genomics, which involve numerous primate species, have proven highly reliable at identifying constrained functional elements.17 Identifying accelerated regions, however, is more complex, requiring statistical approaches that detect positive selection while distinguishing it from rapid sequence evolution simply due to loss of functional constraint. Current approaches to regulatory divergence, including those focused on human accelerated regions (HARs), have sought sequences that are highly conserved across mammals or primates but carry an excess of substitutions on specific branches, such as the human branch since divergence from chimpanzees.20 Another approach, human ancestor quickly evolved regions (HAQERs), addresses sequence divergence from an inferred human-chimpanzee ancestral sequence, specifically in previously unconstrained regions of the genome.21 While HARs have identified many relevant loci, including fetal brain enhancers,22 this approach assumes that human specific adaptations mainly arise in regions that were under strong evolutionary constraint during all of primate evolution. This leads to overlooking neofunctionalization, adaptations emerging from less conserved or previously neutral sequences.23 Additionally, the statistical power to detect these regions is strongly influenced by the number of genomes and taxonomic relationships among species used for detection. Fewer numbers and less diverse species increase the risk of both false positives and false negatives. Previous analyses using few primate genomes identified far fewer HARs than more recent studies using a wider range of primate genomes.24 Moreover, when larger numbers of genomes were considered, hundreds of previously identified HARs no longer showed significant acceleration in some cases because the “human-specific” changes originally identified were no longer unique, also occurring in other primates once more data were included.24 Current methods for identifying positive selection may also depend on biased neutral proxies to set a baseline substitution rate. These neutral proxies include ancient repeats or 4-fold degenerate sites, which may also be subject to cryptic selective pressures or mutational biases.23 Ancient repeats can be co-opted as regulatory elements and remain functionally constrained, while 4-fold degenerate sites may experience selective pressures through different mechanisms including exonic splice regulation, codon usage, mRNA stability, or others.25,26,27,28 Using these biased sequences as neutral proxies can result in false signals of acceleration or mask true adaptive changes.23 Additionally, many prior analyses have identified HARs through genome-wide approaches rather than by targeting regions with known activity in specific tissues and have relied on predictions of regulatory activity inferred from sequence conservation22,29 While this strategy is comprehensive, it does not take into account the biological context in which regulatory elements are active. Sequence conservation is used to identify candidate HARs as well as to argue for their functional importance. Conservation suggests, but may not always necessitate, functional importance; conserved regulatory regions can be active in some species but inactive in others.30,31 Similarly to HARs, HAQERs are initially identified from sequence conservation and divergence with a posteriori experimental validation. Regardless, experimental validation of predicted regions of interest has relied on massively parallel reporter assays, induced pluripotent stem cells (iPSCs), or other in vitro systems, which limits our understanding of function during actual human brain development.32,33,34,35 More importantly for the purposes of understanding evolutionary processes, especially in humans, nearly half of the regulatory active regions in the human genome are not conserved across species.36 In the current study, we used an atlas of experimentally validated brain-active CREs defined by chromatin accessibility (DNase-seq and ATAC-seq), chromatin immunoprecipitation sequencing (ChIP-seq), and histone modifications assays across cortical and subcortical brain regions and from the PsychENCODE Consortium.37 We also leveraged an expanded set of 243 primate genomes currently available to study the evolution of these CREs across the diversity of primates and specifically in the human lineage.17 We applied a branch-specific positive selection test tailored to non-coding elements to detect regulatory regions under directional selection specifically on the human lineage using validated neutral models.23 Using this integrated approach enabled a rigorous and systematic identification of brain-active CREs that have been subject to human-specific selective pressures, potentially underlying uniquely human cognitive and neuropsychiatric traits. Material and methods Brain-active CRE dataset The brain-active CREs (b-CREs) were obtained from the PsychScreen web platform (https://psychscreen.wenglab.org). It contains 361,844 b-CREs identified through ATAC-seq and DNase-seq chromatin accessibility assays integrated with ChIP-seq data for histone modifications