One Cell At a Time (OCAT): a unified framework to integrate and analyze single-cell RNA-seq data

Area of Science:

• Computational Biology
• Bioinformatics
• Genomics

Background:

• Large-scale single-cell RNA sequencing (scRNA-seq) data integration is crucial for aggregating biological insights.
• Existing integration methods struggle with efficiency, especially for multiple large datasets.
• The need for robust methods that handle batch effects without complex preprocessing is significant.

Purpose of the Study:

• To introduce OCAT (One Cell At a Time), a novel machine learning method for efficient integration of multiple large-scale scRNA-seq datasets.
• To demonstrate OCAT's capability in improving cell type clustering, particularly in challenging scenarios.
• To showcase OCAT's utility in facilitating diverse downstream analyses.

Main Methods:

• OCAT employs sparse encoding of single-cell gene expression for data integration.
• The method integrates data from multiple sources without requiring highly variable gene selection.
• OCAT performs integration without explicit batch effect correction.

Main Results:

• OCAT demonstrates efficient integration of multiple large-scale scRNA-seq datasets.
• The method achieves state-of-the-art performance in cell type clustering.
• OCAT shows particular strength in clustering datasets with non-overlapping cell types.

Conclusions:

• OCAT provides an efficient and effective solution for integrating large-scale scRNA-seq data.
• The method advances cell type identification and analysis in complex biological systems.
• OCAT is a valuable tool for a wide range of scRNA-seq downstream analyses.