torbcellselection

To separate T and B cells from mixed scRNA-seq data, you can use clonotype percentages from VDJ data, marker gene expression like CD3 or CD19, custom selector expressions, or automatic k-means clustering to identify and filter specific cell populations.

K-means clustering identifies T cells and non-T cells by automatically grouping mixed scRNA-seq data into clusters based on expression patterns, allowing unsupervised separation without requiring predefined marker genes or VDJ clonotype information.

You do not need VDJ data to isolate B cells if you set ignore_vdj to true, allowing you to rely on marker genes like CD19 and MS4A1 or custom selector expressions instead of scBCR-seq clonotype percentages.

The best way to separate T cells using scTCR-seq data is leveraging clonotype percentage thresholds from VDJ data, which precisely identifies T cell clusters while integrating with SeuratClusteringOfAllCells workflows.

You can use custom selector expressions to filter non-T cells in Seurat, applying your own logical criteria to mixed scRNA-seq data to retain specific clusters and output a filtered Seurat object with detailed selection reports.

T cell separation requires SeuratClusteringOfAllCells data because the selection process validates inputs and performs cluster-level filtering on pre-clustered scRNA-seq data, unless the ignore_vdj parameter is set to true.