Single-Cell Alternative Splicing Analysis with Expedition Reveals Splicing Dynamics during Neuron Differentiation.

Area of Science:

• Molecular Biology
• Genomics
• Computational Biology

Background:

• Alternative splicing (AS) generates significant protein diversity from a limited number of genes.
• Understanding the biological and evolutionary significance of AS variation at the single-cell level remains challenging.

Purpose of the Study:

• To develop a computational framework (Expedition) for detecting and analyzing alternative splicing patterns in single cells.
• To investigate the prevalence and characteristics of AS variation during cellular differentiation.

Main Methods:

• Expedition framework: outrigger (splice graph transversal), anchor (Bayesian modality assignment), bonvoyage (non-negative matrix factorization visualization).
• Application to single pluripotent stem cells during neuronal differentiation.

Main Results:

• Identified that up to 20% of alternative splicing (AS) exons exhibit bimodality in single cells.
• Bimodal exons are associated with conserved intronic sequences, distinct cis-regulatory motifs, and cell-type-specific splicing.
• Observed high dynamism of bimodal exons during cellular transitions, preserving reading frames.

Conclusions:

• Expedition provides a novel approach to characterize AS complexity in single cells.
• Bimodal AS reveals intricate cell states not apparent through conventional gene expression analysis.
• Systematic single-cell AS analysis redefines the understanding of AS complexity in cell biology and evolution.