Brain imaging genomics has opened powerful avenues for understanding the genetic underpinnings of human brain structure and function. Yet, traditional brain atlases, designed purely on neuroanatomical landmarks, have limited the ability to discover meaningful genetic associations. To bridge this gap, researchers have introduced GIANT — the Genetically Informed brAiN aTlas — a groundbreaking resource that simultaneously accounts for both genetic and neuroanatomical variations.
Unlike conventional atlases, GIANT is built by clustering brain voxels based not just on their spatial proximity, but also on their voxel-wise heritability — a measure of how much genetic factors explain variations in brain structure. By integrating these two dimensions, GIANT defines brain regions that are genetically more homogeneous internally and genetically distinct from each other, offering a sharper and more biologically relevant map of the brain.
This innovation holds important implications. Compared to traditional brain atlases, GIANT exhibits smaller intra-region genetic variability and larger inter-region genetic differences, making it a superior tool for studying the genetics of brain morphology. Researchers demonstrated that GIANT’s regions capture higher SNP heritability, show enhanced polygenicity, and lead to polygenic risk scores that explain more variation in brain volumetric traits than those derived from anatomy-based atlases.
The power of GIANT is rooted in robust validation. It was developed using large-scale imaging and genetic data from the UK Biobank and the Alzheimer’s Disease Neuroimaging Initiative, covering over 40,000 subjects across diverse genetic ancestries. Rigorous stability checks, test-retest reliability studies, and homogeneity evaluations confirmed that GIANT’s brain regions are consistent, reliable, and biologically meaningful across populations and conditions.
Beyond mapping the brain, the study also explored the genetic architecture of GIANT-defined regions. Researchers conducted genome-wide association studies (GWAS) to link specific brain regions to genetic variants, uncovering hundreds of associations. Many of these variants were located in noncoding regions of the genome, highlighting the complex regulatory mechanisms behind brain structure. Functional annotation of these variants suggested links to various traits, including brain volume, cognitive abilities, and even disorders like autism spectrum disorder.
Importantly, GIANT also facilitates gene mapping through positional, eQTL, and chromatin interaction strategies, offering new insights into potential therapeutic targets. For instance, brain regions defined by GIANT were linked to genes involved in brain development, neuroinflammation, and synaptic signaling — pathways critical for understanding neurological and psychiatric diseases.
The creation of GIANT marks a pivotal step toward genetically informed brain mapping. It empowers researchers to dissect the genetic basis of brain traits with higher precision, enhances the discovery power of imaging genomics studies, and lays the foundation for future work in personalized medicine and gene-targeted therapies.
While GIANT represents a major advance, the study acknowledges certain limitations — such as the reliance on European-ancestry data and the focus on volumetric imaging features. Future directions include extending the atlas to more diverse populations and integrating other brain phenotypes like cortical thickness and surface area.
Credits : Ramya S, BioCogniz
Reference : Nat Commun 16, 3524 (2025).
