ASPEN: Robust detection of allelic dynamics in single cell RNA-seq

Area of Science:

• Genomics
• Computational Biology
• Molecular Biology

Background:

• Single-cell RNA sequencing (scRNA-seq) of F1 hybrids offers insights into gene regulatory mechanisms.
• Allelic measurements in scRNA-seq are challenged by technical noise and low read counts.
• Accurate modeling of allelic mean and variance is crucial for understanding gene regulation.

Purpose of the Study:

• To introduce ASPEN, a novel statistical method for modeling allelic mean and variance in single-cell transcriptomic data.
• To enhance the sensitivity and accuracy of detecting allelic imbalance and variance changes at the single-cell level.
• To apply ASPEN to biological systems for discovering regulatory patterns.

Main Methods:

• ASPEN integrates a sensitive mapping pipeline with a moderated beta-binomial model.
• Adaptive shrinkage is employed to differentiate allelic imbalance from changes in allelic variance.
• The method was validated using both simulated and empirical scRNA-seq datasets.

Main Results:

• ASPEN demonstrated a ~30% increase in sensitivity for single-cell allelic imbalance detection compared to existing methods.
• The method identified genes exhibiting incomplete X inactivation and random monoallelic expression in mouse brain organoids and T cells.
• Significant deviations in allelic variance were detected, revealing patterns of regulatory control and flexibility.

Conclusions:

• ASPEN provides a robust framework for analyzing allelic data in scRNA-seq, overcoming limitations of technical noise.
• The findings highlight reduced variance in essential genes, suggesting stringent regulatory control.
• Increased variance at neurodevelopmental and immune loci indicates regulatory flexibility, offering new avenues for research.