Detecting Fear-Memory-Related Genes from Neuronal scRNA-seq Data by Diverse Distributions and Bhattacharyya Distance

Area of Science:

• Computational Biology
• Genomics
• Neuroscience

Background:

• Detecting differentially expressed genes (DEGs) is crucial for single-cell RNA sequencing (scRNA-seq) data analysis.
• High cellular heterogeneity and dropout noise in scRNA-seq data challenge the precision and robustness of existing DEG detection methods that rely on single distributions.

Purpose of the Study:

• To develop a more precise and robust method for DEG detection in scRNA-seq data.
• To address the limitations of current DEG analysis tools by incorporating diverse gene expression distributions.

Main Methods:

• Proposed DEGman, a novel computational method utilizing diverse gene expression distributions combined with Bhattacharyya distance.
• DEGman automatically selects optimal distributions and employs permutation testing on Bhattacharyya distances for DEG identification.
• Evaluated DEGman against popular DEG analysis tools using simulated and real scRNA-seq datasets.

Main Results:

• DEGman demonstrated improved sensitivity and precision balance compared to existing methods.
• Application to mouse neuron scRNA-seq data identified known and novel fear-memory-related genes.
• Discovered 25 common DEGs across five neuron clusters, functionally enriched for synaptic vesicles, suggesting their role in memory formation.

Conclusions:

• DEGman offers enhanced performance for DEG analysis in complex scRNA-seq datasets.
• The method's ability to leverage diverse distributions improves accuracy in real-world applications.
• Findings highlight the critical role of synaptic vesicle dynamics in remote memory formation.