Searching large-scale scRNA-seq databases via unbiased cell embedding with Cell BLAST.

Area of Science:

• Genomics
• Computational Biology
• Bioinformatics

Background:

• Single-cell RNA sequencing (scRNA-seq) is crucial for understanding cellular heterogeneity.
• The increasing volume of public scRNA-seq data necessitates efficient cell-querying methods.
• Accurate cell-to-cell similarity measures and robust batch effect handling are vital for data utilization.

Purpose of the Study:

• To develop an accurate and robust cell-querying method for scRNA-seq data analysis.
• To enable effective utilization of existing annotations for newly sequenced cells.
• To provide a comprehensive solution for cell type annotation and differentiation potential assessment.

Main Methods:

• Development of Cell BLAST, a novel cell-querying method.
• Utilizing a neural network-based generative model for data analysis.
• Implementing a customized cell-to-cell similarity metric to address batch effects.

Main Results:

• Cell BLAST demonstrates high accuracy in annotating discrete cell types.
• The method effectively assesses continuous cell differentiation potential.
• Cell BLAST successfully identifies novel cell types within scRNA-seq datasets.
• Extensive benchmarks and case studies validate the tool's performance.

Conclusions:

• Cell BLAST offers an accurate and robust solution for scRNA-seq cell querying and annotation.
• The tool facilitates the curation of newly sequenced cells using public annotations.
• Cell BLAST, integrated with a reference database and web server, provides a user-friendly platform for real-world applications.