Human variation in population-wide gene expression data predicts gene perturbation phenotype

Area of Science:

• Genomics
• Systems Biology
• Bioinformatics

Background:

• Population-scale datasets offer insights into genetic and environmental influences on gene expression.
• Gene expression variance can simulate loss- or gain-of-function experiments within populations.

Purpose of the Study:

• To introduce huva (human variation), a novel approach for inferring gene function and phenotype-expression relationships using population data.
• To demonstrate the utility of huva in identifying the phenotypic relevance of genes and stratifying them by biological function across the human transcriptome.

Main Methods:

• Leveraging multi-layered, population-scale data to derive experimental groups with low and high expression of a gene of interest (GOI).
• Comparing transcriptional profiles and functional parameters between these derived groups.
• Generalizing the approach to analyze approximately 16,000 human genes.

Main Results:

• The huva approach robustly identifies the phenotypic relevance of a GOI.
• Genes can be effectively stratified according to biological functions.
• huva predicted monocytes as key in STAT1 mutation pathophysiology, which was validated in a clinical cohort.

Conclusions:

• huva provides a powerful framework for gene function discovery and phenotype-genotype correlation using population variation.
• The method is scalable to the entire human transcriptome.
• This approach has implications for understanding genetic disorders and identifying cellular mechanisms involved in disease.