DeepKINET: a deep generative model for estimating single-cell RNA splicing and degradation rates

Area of Science:

• Molecular Biology
• Genomics
• Computational Biology

Background:

• Messenger RNA (mRNA) splicing and degradation are crucial for gene expression regulation.
• Abnormalities in these processes are linked to various diseases.
• Existing methods for estimating kinetic rates often assume uniformity across cells, limiting their precision.

Purpose of the Study:

• To develop a novel deep generative model, DeepKINET, for estimating mRNA splicing and degradation rates at single-cell resolution.
• To overcome the limitations of existing methods that assume uniform kinetic rates.
• To enable a deeper understanding of post-transcriptional regulation heterogeneity.

Main Methods:

• DeepKINET, a deep generative model, was developed to analyze single-cell RNA sequencing (scRNA-seq) data.
• The model estimates splicing and degradation rates at the individual cell level.
• Performance was validated using simulated datasets and experimental metabolic labeling data.

Main Results:

• DeepKINET demonstrated superior performance compared to existing methods on both simulated and metabolic labeling datasets.
• Application to forebrain and breast cancer data identified RNA-binding proteins contributing to kinetic rate diversity.
• Analysis of erythroid lineage cells revealed the impact of splicing factor mutations on target genes.

Conclusions:

• DeepKINET provides a powerful tool for dissecting cellular heterogeneity in post-transcriptional regulation.
• The model accurately estimates RNA kinetic rates at single-cell resolution, advancing the study of gene expression.
• DeepKINET facilitates the identification of regulatory factors and the impact of genetic variations on RNA dynamics.