Process single-cell RNA-seq datasets into cleaned, normalized, clustered results.
npx skills add https://github.com/dralkh/seerai --skill scanpy-dralkh
Please help me install this Agent Skill. Skill: scanpy Source: https://github.com/dralkh/seerai/tree/main/skills/scanpy Command: npx skills add https://github.com/dralkh/seerai --skill scanpy-dralkh
This Skill turns raw single-cell RNA-seq data into a clean, interpretable analysis workflow, reducing the time spent on manual QC, preprocessing, clustering, and annotation.
It covers quality control, normalization, highly variable gene selection, dimensionality reduction, Leiden clustering, marker discovery, cell-type annotation, batch correction, pseudobulk aggregation, and publication-ready plotting. It is especially useful for researchers working with AnnData files, 10x Genomics outputs, CSV or loom inputs, and projects that need consistent, reproducible Scanpy workflows with reusable scripts and references.
Use the scanpy skill to analyze my single-cell RNA-seq dataset from raw counts through clustering, marker identification, and annotation.
High-intent search queries and answers about installing and using this skill.
Single-cell RNA-seq analysis transforms raw counts into clustered results through QC filtering, normalization, dimensionality reduction, Leiden clustering, marker discovery, and cell-type annotation for AnnData, 10x, CSV, and loom inputs.
Pseudobulk aggregation sums single-cell RNA-seq counts per cell type and sample to reduce technical noise and enable differential expression analysis. It is used after clustering and cell-type annotation to generate robust statistical results for biological replicates.
Batch correction for AnnData files integrates harmonypy and bbknn to remove technical variations across samples. These tools align cell embeddings from different batches after normalization to produce clean, integrated clustering results.
Yes, UMAP visualization and publication-ready plotting are supported using matplotlib. The workflow generates UMAP embeddings for dimensionality reduction and produces publication-ready plots directly from processed AnnData objects.
Quality control and doublet detection for 10x Genomics outputs filter low-quality cells and remove artifact doublets from raw count matrices. This preprocessing step cleans data before normalization to ensure accurate downstream clustering and annotation.
Highly variable gene selection filters out noisy genes before dimensionality reduction to focus the analysis on biologically meaningful variation. This step reduces computational load and improves the resolution of Leiden clustering and UMAP embeddings.
ya-wayProfile raw 10x Genomics single-cell matrices to output QC distributions as JSON.
TianGzlabPerform pseudobulk differential expression analysis on spatial transcriptomics AnnData objects.
TianGzlabIntegrate multi-sample single-cell RNA sequencing data with Harmony, scVI, Seurat CCA/RPCA, BBKNN, and fastMNN.
pur3v4d3rAnalyze single-cell RNA-seq data with Scanpy for QC, clustering, and annotation.
