sc-cell-annotation

Automated single-cell type annotation assigns biological labels to cells or clusters using marker genes, pre-trained CellTypist models, or reference datasets like SingleR. This Skill supports custom references and marker lists to ensure reproducible biological interpretation of your sequencing data.

Labeling immune cell populations in an AnnData object requires preprocessed data with normalized counts. This Skill uses pre-trained CellTypist immune models and SingleR reference datasets to automatically identify and assign labels to specific immune populations like T cells and B cells.

Single-cell type annotation does not strictly require clustering information but does need preprocessed AnnData objects with normalized counts. Clustering information is optionally supported to help assign biological cell types to distinct cell groups rather than individual cells.

Custom marker gene lists can be used for single-cell annotation instead of pre-trained models. This Skill integrates multiple annotation strategies including marker genes, CellTypist, SingleR, and scmap, allowing consensus-based labeling using your own custom references.

CellTypist uses pre-trained machine learning models for rapid labeling, while SingleR relies on reference dataset correlations for annotation. This Skill integrates both methods alongside marker gene matching, enabling consensus-based cell type labeling for more robust results.