BrainGENIE

BrainGENIE predicts brain-region-specific gene expression from peripheral blood transcriptomes to enable inference of neural gene expression patterns for neurogenetic research.

Key Features:

• Predictive Capability: Predicts brain-tissue-specific expression for between 1,733 and 11,569 genes with false-discovery rate-adjusted p-values less than 0.05, including transcripts not reliably predicted by PrediXcan.
• Validation and Generalizability: Predictions were validated in external datasets, including within-individual data from the Religious Orders Study and Memory and Aging Project, confirming 39% of predictions in the dorsolateral prefrontal cortex and 23% in the caudate nucleus when compared with observed postmortem brain samples.
• Diagnostic Insight: Recapitulates diagnosis-related changes in brain gene expression more accurately than direct blood-to-brain correlations or predictions from PrediXcan.
• Complementary Functionality: Complements existing transcriptome-imputation tools and often outperforms other transcriptome-imputation methods by providing biologically meaningful insights.

Scientific Applications:

• Non-invasive brain expression inference: Enables inference of brain-region gene expression from peripheral blood samples without direct brain tissue sampling.
• Large-scale genetic studies: Facilitates cohort-wide analyses of brain-related gene expression using blood transcriptomes.
• Neurodegenerative disease research: Supports investigation of complex neurological diseases, including Alzheimer's disease, via inferred brain expression signatures.
• Cognitive and neurogenetic studies: Supports exploration of gene expression patterns linked to cognitive function and other neurodegenerative disorders.

Methodology:

Integration of peripheral blood transcriptome data to impute gene expression levels specific to various brain regions.