Linking Genetic Risk to Disease-Relevant Cellular States via Metacell-Informed Modeling with ICePop

Area of Science:

• Genomics
• Computational Biology
• Systems Biology

Background:

• Genome-wide association studies (GWAS) identify genetic loci for complex diseases but struggle to pinpoint cellular mechanisms.
• Existing methods for integrating GWAS with single-cell transcriptomics have limitations in statistical power or resolution.
• A tradeoff exists between cell-type level analysis (high power, low resolution) and single-cell level analysis (high resolution, low power).

Purpose of the Study:

• To introduce ICePop (Informative Cell Populations), a novel framework to bridge the gap between GWAS and single-cell data.
• To achieve high statistical power while resolving heterogeneous disease signals within specific cellular states.
• To enable the generation of cell-state-specific hypotheses for disease mechanisms and therapeutic targets.

Main Methods:

• ICePop performs disease-cell type association analysis at metacell resolution, balancing statistical power and cellular heterogeneity.
• The framework was evaluated using simulations against existing methods like seismic and scDRS.
• ICePop was applied to the Tabula Muris dataset, analyzing 81 traits across 120 cell types.

Main Results:

• ICePop demonstrated superior power in simulations for detecting disease effects in cellular subpopulations.
• The analysis identified 2,178 disease-cell type associations, revealing specific cell vulnerabilities in ulcerative colitis and lung function.
• Clustering diseases by ICePop's metacell association profiles revealed groupings distinct from genetic risk-based clusters.
• Enrichment of genetic risk in specific enteric neuron subtypes was identified for autism spectrum disorder.

Conclusions:

• ICePop effectively resolves the tradeoff between statistical power and cellular heterogeneity in GWAS-based disease association studies.
• The framework provides a powerful tool for identifying disease-relevant cell states within broader cell types.
• ICePop facilitates the development of testable, cell-state-specific hypotheses for complex diseases.