popv-cell-annotation

What problem does it solve?

This Skill automates the complex task of annotating cell types in single-cell RNA sequencing (scRNA-seq) data by leveraging the consensus of multiple algorithms, providing more robust and reliable results than single-method approaches.

Core Features & Use Cases

• Ensemble Annotation: Integrates 10+ algorithms (e.g., KNN-Harmony, scVI, CellTypist, Random Forest) for a consensus cell type prediction.
• Uncertainty Quantification: Provides an agreement score to identify novel or ambiguous cell states where methods disagree.
• Batch Effect Robustness: Designed to handle substantial batch effects between reference and query datasets.
• Use Case: Annotate a new single-cell dataset from a patient's tumor sample by transferring cell type labels from a large, well-curated reference atlas, while also identifying rare cell populations that might be missed by a single annotation tool.

Quick Start

Use the popv-cell-annotation skill to annotate your query dataset using a labeled reference atlas.