CCmed: cross-condition mediation analysis for identifying replicable trans-associations mediated by cis-gene...

Area of Science:

• Genetics
• Bioinformatics
• Systems Biology

Background:

• Trans-acting expression quantitative trait loci (eQTLs) contribute significantly to gene expression variation but are difficult to detect due to weak individual effects.
• Genetic influences on distal genes are often mediated by the expression of nearby (cis) genes.
• Detecting trans-associations requires linking single nucleotide polymorphisms (SNPs) to cis-gene expression and subsequent gene-gene correlations.

Purpose of the Study:

• To develop a robust method for identifying cis-mediated trans-associations with replicable effects across conditions.
• To enhance the detection of genetic effects on gene expression that are shared across tissues and studies.
• To identify trans-genes associated with complex diseases like schizophrenia.

Main Methods:

• Proposed Cross-Condition Mediation analysis (CCmed) integrating cis-association and gene-gene conditional correlation statistics from multiple tissues/studies.
• Developed CCmedmost and CCmedspec variations to detect cross-tissue and tissue-specific trans-associations, respectively.
• Analyzed 13 brain tissues from the Genotype-Tissue Expression (GTEx) project.

Main Results:

• Identified numerous cis-mediated trans-associations across brain tissues, with many validated in replication studies.
• Discovered trans-genes linked to schizophrenia loci in at least two brain tissues.
• CCmed software is publicly available for broader research application.

Conclusions:

• CCmed effectively detects replicable cis-mediated trans-associations across diverse conditions and tissues.
• The method facilitates the discovery of genetic regulatory mechanisms underlying complex traits and diseases.
• This approach advances the understanding of how genetic variation influences gene expression networks.