How does the k-means clustering method determine which cells to select?

It calculates the average expression of indicator genes and clonotype percentages per cluster, performs k-means (K=2) clustering, and then selects the cluster with the higher signature of the target cell type.

What are the required input files for this skill?

It requires a Seurat object (usually from a clustering process) and, unless ignore_vdj is enabled, an immune repertoire data file containing VDJ information.

Which marker genes are used by default for T cell selection?

The skill defaults to using the CD3E marker, but it can be configured to use a combination of CD3E, CD3D, and CD3G for more robust T cell identification.

Can I use this skill if I don't have VDJ data?

Yes, by setting ignore_vdj = true in the configuration, the skill will rely solely on marker gene expression and k-means clustering to identify and select the target cell types.

T and B Cell Selection Tool

Name: T and B Cell Selection Tool
Author: pwwang

bypwwang

•

Ciencia de Datos y ML

Automates the isolation of T and B cell populations from mixed single-cell datasets using marker gene expression and VDJ clonotype data.

This skill facilitates the precise separation of T or B cells from mixed single-cell populations, a critical step in immunology workflows. It leverages multiple data streams, including VDJ clonotype percentages and canonical marker gene expression (such as CD3 for T cells or CD19/CD20 for B cells), providing flexible selection methods via automated k-means clustering or custom R-based selector expressions. Designed to work within the Seurat-based immunopipe framework, it ensures that downstream TCR/BCR-specific analyses are performed on high-purity cell populations.

Características Principales

0120 GitHub stars

02Pre-configured marker sets for T cell (CD3) and B cell (CD19/CD20) lineages

03Multi-modal selection using combined scRNA-seq and scTCR/BCR-seq (VDJ) data

04Automated k-means clustering for unsupervised, data-driven cell type identification

05Support for custom R selector expressions for advanced filtering logic

06Seamless integration with Seurat objects and immunopipe downstream processes

Casos de Uso

01Identifying B cell clusters in peripheral blood mononuclear cells (PBMCs) when VDJ data is missing

02Filtering for high-confidence clonal populations using combined gene expression and clonotype thresholds

03Isolating T cells from a whole-tissue single-cell dataset before performing TCR repertoire analysis

What are Skills?·How to Install

Install with 🐟 Skill.Fish

npx skillfish add pwwang/immunopipe torbcellselection

For use in Claude.ai and ChatGPT

Download Skill