SCMarker

SCMarker is a computational tool for unsupervised marker gene selection in single-cell RNA-sequencing (scRNA-seq) data analysis. It aims to identify informative genes that can discriminate between different cell subpopulations and reduce the dimensionality of the data, thereby facilitating accurate cell typing.

SCMarker employs ab initio, meaning it does not rely on prior knowledge or external reference data. Instead, it identifies marker genes based on two key criteria: 1) genes that exhibit subpopulation-discriminative expression levels, and 2) genes that are co-expressed or mutually exclusively expressed with other genes.

By applying SCMarker to various scRNA-seq datasets from multiple tissue types and coupling it with different clustering algorithms, the tool consistently improves cell-type classification accuracy and selects biologically meaningful marker genes.