scMAGS

scMAGS selects cell-type marker genes from single-cell RNA sequencing (scRNA-seq) data to support spatial transcriptomics by identifying compact marker panels that distinguish cell types.

Key Features:

• Marker Gene Selection: Identifies a limited set of genes that distinguish different cell types from scRNA-seq data.
• Filtering Step: Applies a preselection filter to identify candidate genes for downstream marker selection.
• Cluster Validity Indices: Evaluates marker sets using indices such as the Silhouette index and the Calinski-Harabasz index.
• Scalability and Efficiency: Demonstrates scalability to large datasets and reduced memory usage for handling millions of cells.

Scientific Applications:

• Spatial Transcriptomics: Enables selection of marker genes for mapping scRNA-seq-defined cell types to spatial transcriptomics data.
• Cell-type Identification: Supports accurate identification and discrimination of cell types using compact marker panels.
• Gene Regulatory Analysis: Facilitates uncovering regulatory relationships between genes at the single-cell level.
• Research Domains: Applicable to developmental biology, oncology, and regenerative medicine where spatially resolved gene expression is informative.

Methodology:

Performs candidate-gene filtering followed by marker-gene selection and evaluates marker sets using cluster validity indices such as the Silhouette index and the Calinski-Harabasz index.