scPLAN: a hierarchical computational framework for single transcriptomics data annotation, integration and cell-type...

Area of Science:

• Computational biology
• Genomics
• Bioinformatics

Background:

• Single-cell RNA sequencing (scRNA-seq) is vital for transcriptomic profiling.
• Accurate cell-type identification and data integration are critical for scRNA-seq analysis.
• Existing methods struggle with hierarchical cell organization and integrating datasets of varying annotation depths.

Purpose of the Study:

• To introduce scPLAN, a hierarchical computational framework for scRNA-seq data analysis.
• To address limitations in current cell-type annotation and data integration methods.
• To enable consistent refinement of cell-type labels across datasets with different annotation resolutions.

Main Methods:

• Developed scPLAN, a hierarchical computational framework.
• Utilized a reference dataset structured with a hierarchical cell-type tree for annotation.
• Implemented layer-by-layer identification of potential novel cell types.
• Designed scPLAN to integrate scRNA-seq datasets with varying annotation depths.

Main Results:

• scPLAN effectively annotates unlabeled scRNA-seq data.
• Identified potential novel cell types systematically.
• Successfully integrated datasets with diverse cell-type label resolutions.
• Demonstrated consistent refinement of cell-type labels for lower-resolution datasets.

Conclusions:

• scPLAN provides a robust framework for scRNA-seq data analysis.
• The hierarchical approach enhances cell-type annotation accuracy and consistency.
• scPLAN facilitates the integration of heterogeneous scRNA-seq datasets.
• The framework aids in discovering novel cell types and refining existing annotations.