scPrediXcan integrates deep learning methods and single-cell data into a cell-type-specific transcriptome-wide...

Area of Science:

• Genomics
• Computational Biology
• Systems Biology

Background:

• Transcriptome-wide association studies (TWASs) identify disease-associated genes but lack cellular resolution.
• Limited sample sizes and sparse cell-type-specific expression data hinder mechanistic understanding.
• Existing TWAS methods struggle to capture complex gene-regulatory relationships.

Purpose of the Study:

• To introduce scPrediXcan, a novel framework integrating deep learning with TWAS.
• To improve the identification of disease-causal genes and their cellular mechanisms.
• To enhance the understanding of complex diseases like type 2 diabetes and systemic lupus erythematosus.

Main Methods:

• scPrediXcan utilizes deep learning (ctPred) to predict cell-type-specific gene expression from DNA sequences.
• It integrates these predictions into the established TWAS framework.
• The approach was applied to genetic data for type 2 diabetes and systemic lupus erythematosus.

Main Results:

• scPrediXcan identified more candidate causal genes compared to canonical TWAS.
• The method explained a greater proportion of genome-wide association study (GWAS) loci.
• scPrediXcan provided novel insights into the cellular specificity of TWAS findings.

Conclusions:

• scPrediXcan significantly advances the ability to pinpoint disease mechanisms at the cellular level.
• The framework offers a powerful tool for dissecting the genetic architecture of complex diseases.
• This approach promises to deepen our understanding of disease etiology and cellular function.