Variant to function mapping at single-cell resolution through network propagation.

Area of Science:

• Genomics
• Computational Biology
• Systems Biology

Background:

• Genome-wide association studies (GWAS) identify genetic variants linked to diseases, but pinpointing causal mechanisms is challenging.
• Analyzing single-cell epigenomic data for disease-relevant cell types, states, and trajectories is often hindered by data sparsity and noise.

Purpose of the Study:

• To present SCAVENGE, a novel computational algorithm for mapping disease-causal genetic variants to their specific cellular contexts at single-cell resolution.
• To overcome limitations in identifying disease-relevant cell types and states from noisy single-cell epigenomic data.

Main Methods:

• SCAVENGE utilizes network propagation to link genetic variants with cellular information.
• The algorithm is applied to single-cell genomic and epigenomic datasets.

Main Results:

• SCAVENGE successfully mapped causal variants to relevant cellular contexts across different biological systems.
• The method identified key biological mechanisms underlying human genetic variation in blood traits, COVID-19 risk, and acute leukemia predisposition.
• Demonstrated utility in analyzing distinct stages of human hematopoiesis and monocyte/lymphocyte subsets.

Conclusions:

• SCAVENGE provides a robust framework for variant-to-function insights at single-cell resolution.
• The approach offers a general strategy to maximize inferences from single-cell genomic data for understanding genetic variation and disease mechanisms.