Distributed eQTL analysis with auxiliary information

Area of Science:

• Genetics
• Statistical Genomics
• Bioinformatics

Background:

• Expression quantitative trait locus (eQTL) analysis links genetic variations to gene expression levels.
• Existing methods often analyze tissues independently or focus on shared eQTLs, neglecting valuable auxiliary tissue information.
• There is a need for methods that improve eQTL detection in a target tissue by utilizing data from related tissues.

Purpose of the Study:

• To develop a novel statistical framework for enhanced eQTL detection in a target tissue.
• To effectively integrate information from auxiliary tissues to boost statistical power.
• To provide efficient computational approaches for large-scale multi-tissue eQTL analysis.

Main Methods:

• Proposed a statistical framework incorporating shared and specific effects across multiple tissues.
• Developed data-driven and distributed computing strategies for efficient implementation.
• Applied the method to simulated data and real-world GTEx project data.

Main Results:

• The novel framework significantly improves the power of eQTL detection compared to existing methods.
• Simulation studies confirmed the method's efficacy in identifying true eQTLs.
• Real data analysis revealed new insights into tissue-specific and shared eQTLs within the GTEx dataset.

Conclusions:

• The proposed method offers a powerful approach for eQTL discovery by leveraging multi-tissue information.
• Efficient implementation strategies enable its application to large genomic datasets.
• This framework advances our understanding of the genetic architecture of gene expression across diverse human tissues.