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

Area of Science:

• Genetics
• Bioinformatics
• Computational Biology

Background:

• Genome-wide association studies (GWAS) have identified numerous disease-associated genetic variants over two decades.
• However, pinpointing specific causal genes from GWAS data remains challenging due to complex data structures and potential biases.
• Mendelian Randomization (MR) using expression quantitative loci (eQTLs) offers a promising approach to infer causality, but existing methods face limitations.

Purpose of the Study:

• To develop a robust whole-genome regulatory network analysis tool (HORNET) for identifying causal genes using summary-level GWAS data.
• To address biases and improve inference accuracy in MR analyses of eQTL GWAS data.
• To apply HORNET to identify causal genes for schizophrenia across different brain tissues.

Main Methods:

• Development of HORNET, a comprehensive suite of statistical and computational tools.
• Utilizing summary-level GWAS data and eQTL information within a Mendelian Randomization framework.
• Genome-wide analysis of schizophrenia data to assess gene expression causality.

Main Results:

• HORNET provides a robust method for genome-wide causal gene discovery from GWAS data.
• Application of HORNET to schizophrenia revealed differential magnitudes of gene expression causality.
• Causal gene expression differences were observed across various brain tissues in schizophrenia patients.

Conclusions:

• HORNET enhances the ability to identify causal genes from complex GWAS and eQTL data.
• The tool offers improved robustness against biases inherent in existing MR approaches.
• Findings highlight tissue-specific gene expression causality in schizophrenia, paving the way for further research.