Evaluating diabetes and hypertension disease causality using mouse phenotypes.

Area of Science:

• Genetics
• Systems Biology
• Computational Biology

Background:

• Genome-wide association studies (GWAS) identify numerous single nucleotide polymorphisms (SNPs) linked to common diseases.
• Determining the causal genes implicated by these SNPs remains a significant challenge, especially for complex diseases.
• Direct experimental validation of gene causality in mammalian models is often infeasible.

Purpose of the Study:

• To develop a computational method for predicting gene causality in complex diseases using mouse phenotype data.
• To quantify the probability that perturbing a specific gene leads to disease-relevant phenotypes.
• To apply and validate the method using type II diabetes (T2D) and hypertension (HT) as case studies.

Main Methods:

• Leveraged extensive mouse phenotype data.
• Developed a probabilistic method to link gene perturbations to disease phenotypes.
• Analyzed interactome networks and pathway enrichment for high-probability causal genes.

Main Results:

• Identified genes with high probability of causing T2D and HT phenotypes are network hubs.
• These high-probability genes are enriched in signaling pathways regulating metabolism, not metabolic pathways themselves.
• This contrasts with expression data, where genes within metabolic pathways show significant changes.

Conclusions:

• Mouse phenotype-based prediction offers increased coverage and high specificity compared to human genetic data.
• The calculated disease phenotype probabilities can assess the causal likelihood of disease-associated genes and nearby SNPs.
• This approach aids in prioritizing genes for further investigation in complex disease research.