omicverse-single-cell-preprocessing

Convert OmicVerse single-cell AnnData objects into cluster-ready preprocessing and marker-discovery workflows.

13|2|Updated Apr 5, 2026
One-click install
npx skills add https://github.com/omicverse/omicverse-skills --skill omicverse-single-cell-preprocessing-omicverse
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: omicverse-single-cell-preprocessing
Source: https://github.com/omicverse/omicverse-skills/tree/main/src/omicverse_skills/skills/single-cell-preprocessing
Command: npx skills add https://github.com/omicverse/omicverse-skills --skill omicverse-single-cell-preprocessing-omicverse

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill includes references (resource) and assets (resource) components.

What problem does it solve?

This Skill turns raw or partially processed OmicVerse single-cell AnnData objects into a reliable analysis spine, so users do not have to remember which preprocessing, clustering, and marker-discovery steps go together or which keys should exist at each stage.

Core Features & Use Cases

  • End-to-end preprocessing: Handles QC, normalization, HVG selection, scaling, PCA, neighbor graph construction, and embedding generation.
  • Cluster-ready analysis: Runs Leiden clustering after the graph is built and validates the expected AnnData fields before continuing.
  • Marker discovery workflow: Supports both fast raw-count marker discovery with cosg and statistical marker testing with wilcoxon and related methods.
  • Use case: A researcher with a filtered scRNA-seq AnnData object can ask for preprocessing and marker extraction in one guided workflow instead of manually chaining OmicVerse calls and checking storage keys by hand.

Quick Start

Ask the skill to preprocess your OmicVerse single-cell AnnData object, build the graph and Leiden clusters, and then extract and summarize marker genes for each cluster.

Frequently Asked Questions about omicverse-single-cell-preprocessing

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I preprocess raw scRNA-seq data for Leiden clustering and marker discovery?

Single-cell preprocessing applies QC, normalization, HVG selection, scaling, PCA, and neighbor graph construction to AnnData objects, producing cluster-ready output for Leiden clustering and subsequent marker gene extraction.

What is the difference between raw-count and log-normalized marker discovery in single-cell analysis?

Marker discovery for single-cell data supports fast raw-count marker identification using cosg and statistical marker testing using wilcoxon for log-normalized data, requiring explicit branch selection and correct method choice based on your AnnData object's state.

Can I use Scanpy with AnnData for single-cell preprocessing and embedding generation?

Yes, single-cell preprocessing workflows accept AnnData objects and integrate with Scanpy-compatible steps, applying QC, normalization, HVG selection, PCA, neighbor graph construction, and embedding generation to prepare data for downstream Leiden clustering.

How do I validate AnnData keys before running Leiden clustering on single-cell data?

The preprocessing workflow validates expected AnnData fields at each stage before continuing to Leiden clustering, ensuring that required keys for QC, normalization, PCA, and neighbor graph construction exist and are correctly populated.

What's the best way to run end-to-end single-cell preprocessing on filtered AnnData objects?

For filtered single-cell AnnData objects, a guided preprocessing workflow chains normalization, HVG selection, scaling, PCA, neighbor graph construction, embedding, Leiden clustering, and marker extraction, eliminating manual OmicVerse call chaining and key checking.