Comprehensive functional genomic resource and integrative model for the human brain.

Area of Science:

• Neuroscience
• Genetics
• Computational Biology

Background:

• Genetic risk factors for psychiatric disorders are known, but their underlying molecular mechanisms are not fully understood.
• Understanding gene regulation in the brain is crucial for deciphering complex psychiatric conditions.

Purpose of the Study:

• To create a comprehensive online resource for the adult human brain to investigate molecular mechanisms of psychiatric disorders.
• To integrate multi-omics data to build gene regulatory networks and improve psychiatric disease prediction.

Main Methods:

• Generated a large-scale dataset from 1866 individuals, including brain-active enhancers, Hi-C data, and single-cell expression profiles.
• Performed expression quantitative-trait loci (eQTL) analysis for gene expression, chromatin, splicing, and cell-type proportions.
• Integrated diverse datasets to reconstruct gene expression variation and built a gene regulatory network.

Main Results:

• The PsychENCODE resource includes ~79,000 brain-active enhancers and detailed molecular profiles.
• Cell-type proportions explain over 88% of expression variation across populations.
• A gene regulatory network linked genome-wide association study (GWAS) variants to genes, with 321 genes identified for schizophrenia.
• An interpretable deep-learning model using this network improved psychiatric disorder prediction by six-fold compared to polygenic risk scores.

Conclusions:

• The PsychENCODE resource provides a powerful platform for dissecting the molecular basis of psychiatric disorders.
• Integrating multi-omics data and advanced computational models can significantly enhance our understanding of disease etiology and prediction.
• This work identifies key genes and pathways implicated in psychiatric disorders, paving the way for future therapeutic strategies.