PCLDA: An interpretable cell annotation tool for single-cell RNA-sequencing data based on simple statistical methods.

Area of Science:

• Genomics
• Computational Biology
• Bioinformatics

Background:

• Single-cell RNA sequencing (scRNA-seq) provides high-resolution cellular heterogeneity insights.
• Accurate cell-type annotation is critical but challenging due to tool complexity and dataset variability.
• Existing automated annotation tools often rely on complex models, limiting reliability across different experimental protocols.

Purpose of the Study:

• To develop a robust and interpretable cell-type annotation pipeline for scRNA-seq data.
• To address the limitations of complex modeling assumptions in current annotation tools.
• To provide a computationally efficient and accurate alternative for scRNA-seq analysis.

Main Methods:

• Proposed PCLDA pipeline: t-test-based gene screening, Principal Component Analysis (PCA), and Linear Discriminant Analysis (LDA).
• Incorporated novel enhancements within the PCA module to improve performance and robustness.
• Utilized simple, interpretable statistical methods throughout the pipeline.

Main Results:

• PCLDA achieved top-tier accuracy across 22 scRNA-seq datasets and 35 scenarios, outperforming nine state-of-the-art methods.
• Demonstrated consistent performance in both intra-dataset (cross-validation) and inter-dataset (cross-platform) evaluations.
• Showcased stability and superior performance over complex methods when analyzing data from different protocols.
• Highlighted strong interpretability due to linear PCA and LDA modules, enabling direct gene contribution analysis.

Conclusions:

• PCLDA offers a practical and reliable solution for scRNA-seq cell-type annotation.
• The pipeline's simplicity, interpretability, and high performance make it a valuable tool for researchers.
• Enhanced simple statistical methods can effectively address complex challenges in single-cell data analysis.