CIDER

CIDER performs integrative clustering and evaluation of single-cell RNA sequencing (scRNA-Seq) datasets using a meta-clustering framework to mitigate batch effects and assess biological correctness.

Key Features:

• Meta-Clustering Workflow: Employs a meta-clustering approach that leverages inter-group similarity measures to cluster scRNA-Seq data from multiple sources.
• Batch Effect Mitigation: Minimizes batch effects without requiring stringent assumptions about cell population composition across samples.
• Biological Correctness Assessment: Evaluates the biological correctness of integrated datasets and the preservation of biological signals without prior cellular annotations.
• Performance Superiority: Demonstrates improved clustering and integration performance on simulated and real-world datasets with high variability and batch effects.

Scientific Applications:

• Comprehensive Single-Cell Analysis: Enables integrative analysis across multiple scRNA-Seq experiments to reveal cellular heterogeneity and function.
• Robust Data Integration: Supports integration of datasets with varying batch effects and biological variability for large-scale comparative studies.
• Validation of Biological Findings: Assesses whether integrated scRNA-Seq results preserve biologically meaningful signals versus technical artifacts.

Methodology:

Uses a meta-clustering framework that applies inter-group similarity measures to cluster groups of cells rather than individual cells, reducing the impact of batch effects and enabling assessment of biological correctness without prior annotations.