SCMarker

SCMarker selects cell-type–discriminative marker genes from single-cell RNA sequencing (scRNA-seq) data using an unsupervised approach based on expression patterns.

Key Features:

• Ab Initio Unsupervised Marker Selection: Identifies subpopulation-discriminative genes exhibiting co-expression or mutual exclusivity patterns to distinguish cell types without prior labels.
• Technology-Agnostic scRNA-seq Analysis: Applies to scRNA-seq datasets generated by platforms such as Drop-seq and SMART-seq and improves downstream clustering and cell-type classification.

Scientific Applications:

• Cell-Type Identification in Heterogeneous Tissues: Enhances clustering accuracy and marker discovery for analyzing cellular heterogeneity in complex tissues.

Methodology:

SCMarker performs unsupervised gene selection by detecting genes with subpopulation-specific expression distributions and identifying co-expressed or mutually exclusive marker sets, which are then used to improve clustering and classification of scRNA-seq data.