An expression-directed linear mixed model discovering low-effect genetic variants

Area of Science:

• Genetics
• Bioinformatics
• Computational Biology

Background:

• Detecting low-effect genetic variants is crucial for understanding disease pathology and heritability.
• Moderate sample sizes limit the discovery of subtle genetic signals.
• Current methods struggle to effectively estimate the polygenic component in genetic models.

Purpose of the Study:

• To develop a novel method for detecting low-effect genetic variants using moderate sample sizes.
• To improve the estimation of heritability by incorporating gene expression relevance.
• To advance precision medicine through enhanced genetic variant detection.

Main Methods:

• Utilized informative weights from genetically predicted gene expression models.
• Developed an expression-directed linear mixed model (EDLMM) to estimate the polygenic term.
• Incorporated gene expression relevance into the genetic background estimation within the linear mixed model.

Main Results:

• The expression-directed linear mixed model successfully detected subtle signals of low-effect variants in cohorts of ~5,000 individuals.
• Demonstrated significant power gain at the low-effect end of the genetic etiology spectrum for both binary (WTCCC) and quantitative (NFBC1966) traits.
• Substantially improved the estimation of missing heritability by identifying additional low-effect variants.

Conclusions:

• The expression-directed linear mixed model is effective for discovering low-effect genetic variants with moderate sample sizes.
• This approach enhances the estimation of heritability and advances the field of precision medicine.
• Accurate detection of low-effect genetic variants contributes to a better understanding of human diseases.