A web server for comparative analysis of single-cell RNA-seq data.

Area of Science:

• Computational biology
• Genomics
• Bioinformatics

Background:

• Single cell RNA sequencing (scRNA-seq) enables high-resolution profiling of cellular heterogeneity.
• Current cell characterization methods rely on unsupervised analysis and limited marker genes, restricting identification to well-known cell types.

Purpose of the Study:

• To develop an automated pipeline for large-scale, supervised cell type annotation of scRNA-seq datasets.
• To enhance the accuracy and efficiency of cell type identification in complex biological samples.

Main Methods:

• An automated pipeline was created to download, process, and annotate public scRNA-seq data.
• Supervised neural networks were extended for efficient and accurate scRNA-seq data representation.
• The pipeline was applied to over 500 studies, analyzing more than 300 unique cell types.

Main Results:

• Supervised methods demonstrated superior performance in cell type identification compared to unsupervised approaches.
• The pipeline successfully characterized a large number of cell types across diverse scRNA-seq studies.
• A case study illustrated the utility for comparing cell type distributions in healthy versus diseased states.

Conclusions:

• The developed automated pipeline and supervised neural network approach significantly advance large-scale cell type characterization from scRNA-seq data.
• scQuery, a web server, provides an accessible tool for cell type and key gene identification, facilitating biological discovery.