Connectivity Network Feature Sharing in Single-Cell RNA Sequencing Data Identifies Rare Cells.

Area of Science:

• Genomics
• Computational Biology
• Cell Biology

Background:

• Single-cell RNA sequencing (scRNA-seq) is crucial for cell subtype identification.
• Detecting rare cell populations in scRNA-seq data is challenging due to low abundance and subtle expression patterns.
• Existing methods often fail to identify these rare cells effectively.

Purpose of the Study:

• To develop a novel computational approach for robustly identifying rare cells in scRNA-seq data.
• To improve the sensitivity and accuracy of rare cell detection compared to conventional methods.
• To leverage network-based feature sharing for enhanced rare cell identification.

Main Methods:

• Developed SCLCNF (Local Connectivity Network Feature Sharing), a novel algorithm for scRNA-seq data analysis.
• Constructed a cellular connectivity network to analyze cell-to-neighbor relationships.
• Utilized unique feature expression patterns and a rarity score to identify rare cells.

Main Results:

• SCLCNF demonstrated superior performance in detecting rare cells compared to existing techniques.
• The method showed enhanced robustness in identifying rare cell populations.
• Successfully applied to human gastrula datasets for precise rare cell pinpointing.
• Uncovered previously unidentified rare cell populations in sepsis datasets.

Conclusions:

• SCLCNF offers a significant advancement in rare cell identification from scRNA-seq data.
• The network-based approach effectively captures subtle coexpression patterns indicative of rare cells.
• This method has broad applicability in biological and clinical research for discovering rare cell subtypes.