E-MAGMA

E-MAGMA integrates tissue-specific eQTL information with GWAS summary statistics to assign risk variants to putative target genes and compute gene-level association statistics for identifying candidate causal genes underlying complex traits and diseases.

Key Features:

• eQTL-guided variant-to-gene assignment: Assigns risk variants to putative target genes using tissue-specific expression quantitative trait loci (eQTL) information.
• Gene-level statistic computation: Converts GWAS summary statistics into gene-level association statistics to prioritize candidate causal genes.
• Simulation-based performance evaluation: Demonstrated superior performance in simulations using simulated phenotype data with an eQTL reference from chromosome 1 and varying proportions of variance explained by eQTLs (1%, 2%, and 5%) across scenarios with different numbers of causal genes.
• Application to real GWAS datasets: When applied to GWAS summary statistics for five neuropsychiatric disorders, identified a greater number of putative candidate causal genes compared with other eQTL-informed gene-based approaches.

Scientific Applications:

• Causal gene prioritization: Pinpoints genes likely to mediate GWAS signals by integrating tissue-specific eQTLs with summary-level association data.
• Neuropsychiatric genetics: Applied to GWAS of five neuropsychiatric disorders to expand identification of putative causal genes implicated in disease biology.

Methodology:

Inputs are GWAS summary statistics; risk variants are assigned to potential target genes using tissue-specific eQTL information; those associations are converted into gene-level statistics; performance was evaluated using simulations with eQTL reference data from chromosome 1 varying the proportion of variance explained by eQTLs (1%, 2%, 5%) and the number of causal genes, and the method was applied to GWAS summary statistics for five neuropsychiatric disorders.