Predicting genotype-specific gene regulatory networks.

Area of Science:

• Genomics
• Systems Biology
• Bioinformatics

Background:

• Individual genetic variation significantly impacts gene regulation, influencing health and disease.
• Current gene regulatory network construction often overlooks genotype-specific effects, particularly regulatory genetic variants.
• Understanding genotype-phenotype links requires methods that incorporate individual genetic differences.

Purpose of the Study:

• To develop a computational method, EGRET, for inferring genotype-specific gene regulatory networks.
• To identify how individual genetic variation shapes gene regulation within populations.
• To explore genotype-associated regulatory differences relevant to human health and disease.

Main Methods:

• EGRET constructs a genotype-informed prior network using TF motif predictions, eQTL data, and variant effects on TF binding.
• Message passing integrates the prior network with gene expression and TF protein-protein interaction data.
• Validation employed allele-specific expression, chromatin accessibility QTLs, and ChIP-seq data.

Main Results:

• EGRET successfully inferred genotype-specific gene regulatory networks for cell lines and multiple individuals.
• Identified genotype-associated, cell line-specific regulatory differences validated experimentally.
• Discovered cell type-specific regulatory differences linked to disease associations in a large cohort.

Conclusions:

• EGRET is the first method to infer gene regulatory networks reflecting individual genetic variation.
• The approach provides insights into genetic regulatory associations underlying complex phenotypes.
• This method has potential for refining genotype-specific disease risk assessment and treatment strategies.