Cell state-dependent allelic effects and contextual Mendelian randomization analysis for human brain phenotypes.

Area of Science:

• Neurogenetics
• Systems Biology
• Genomics

Background:

• Gene expression quantitative trait loci (eQTLs) are crucial for linking genes to central nervous system (CNS) phenotypes.
• The influence of brain disease on these gene-trait associations remains largely unknown.
• Understanding these effects is vital for accurate genetic interpretation in neurological disorders.

Purpose of the Study:

• To investigate how brain disease affects gene expression and its correlation with genetic variation.
• To determine the reliability of eQTLs in disease versus non-diseased brain tissue.
• To identify robust gene-trait associations and potential biomarkers for CNS outcomes.

Main Methods:

• Analysis of 2,348,438 single-nuclei profiles from 391 human brains (both disease cases and controls).
• Identification of genes with expression correlated to genetic variation, assessing disease-dependent allelic effects.
• Colocalization analysis of genetic variants with 30 CNS traits.
• Application of single-cell Mendelian randomization in non-diseased brains.
• Replication of findings in the UK Biobank.

Main Results:

• 13,939 genes showed expression correlated with genetic variation; 16.7-40.8% exhibited disease-dependent allelic effects.
• 23.6% of 501 gene-trait colocalizations showed disease dependency.
• Analysis of non-diseased brains (n=183) revealed 91 additional colocalizations missed in the mixed cohort.
• Single-cell Mendelian randomization in control brains identified 140 putatively causal gene-trait associations.
• 11 of these associations were replicated in the UK Biobank.

Conclusions:

• Brain disease significantly confounds genetic inferences of gene-phenotype relationships.
• Using non-diseased brain data enhances the accuracy of eQTL interpretation and variant association.
• This approach successfully prioritizes candidate peripheral biomarkers for CNS outcomes.