HORNET: tools to find genes with causal evidence and their regulatory networks using eQTLs

Area of Science:

• Genetics and Genomics
• Statistical Genetics
• Computational Biology

Background:

• Genome-wide association studies (GWAS) have identified numerous disease-associated variants but few causal genes.
• Mendelian randomization (MR) using expression quantitative loci (eQTLs) shows promise for causal gene detection.
• Existing MR methods struggle with complex eQTL GWAS data, leading to bias and incorrect inference.

Purpose of the Study:

• To develop a robust whole-genome regulatory network analysis tool (HORNET) for identifying causal genes.
• To address biases and limitations in current MR approaches for eQTL GWAS data.
• To enable genome-wide causal gene searches using summary-level GWAS data.

Main Methods:

• Development of HORNET, a suite of statistical and computational tools.
• Utilizes summary-level GWAS data to perform genome-wide searches for causal genes.
• Designed to be robust against multiple sources of bias inherent in eQTL GWAS data.

Main Results:

• HORNET successfully performs genome-wide causal gene searches using summary-level GWAS data.
• Application to schizophrenia revealed widespread eQTL effects in the cerebellum.
• In the cortex, eQTL effects were observed to be more localized to specific genetic loci.

Conclusions:

• HORNET provides a powerful and robust method for identifying causal genes from GWAS data.
• The tool overcomes limitations of existing MR approaches, improving causal inference.
• Demonstrates differential regulatory patterns in brain regions for complex traits like schizophrenia.