skills.rest
omicverse
omicverseOFFICIAL
Explore all skills by this author →
WebsiteGitHub

single-popv-annotation

Annotate scRNA-seq query data with consensus cell-type labels via majority voting.

32|5|Updated Mar 12, 2026
Search Tags:#ensemble#scrna-seq#cell-type-annotation#pretrained-models#scvi#majority-voting
One-click install
npx skills add https://github.com/omicverse/omicclaw --skill single-popv-annotation-omicverse
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: single-popv-annotation
Source: https://github.com/omicverse/omicclaw/tree/main/src/omicverse_skills/skills/single-popv-annotation
Command: npx skills add https://github.com/omicverse/omicclaw --skill single-popv-annotation-omicverse

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

PopV addresses inconsistent or fragile single-method cell type annotations by executing multiple classification algorithms and combining their outputs into a consensus label, reducing method-specific failures and increasing annotation robustness for population-level scRNA-seq analyses.

Core Features & Use Cases

  • Ensemble Annotation: Runs up to 10 algorithms (scVI/SCANVI, CellTypist, OnClass, RF, SVM, XGBoost, BBKNN, Harmony, Scanorama) and aggregates results with majority voting.
  • Ontology-aware Consensus: Supports Cell Ontology (CL) hierarchy to resolve labels and produce ontology-refined predictions.
  • Flexible Modes & Hub Models: Offers retrain, inference, and fast modes plus pretrained HubModel integration for large atlases and quick annotation.
  • Visualization & Validation: Provides agreement plots, prediction score summaries, and defensive input validation to guide troubleshooting and method selection.
  • Use Cases: Annotate PBMCs against a reference atlas, project new tissue datasets using pretrained hub models, or run CPU-only classical methods for constrained environments.

Quick Start

Run PopV to annotate your query scRNA-seq dataset against a labeled reference and save consensus cell-type labels and agreement plots to results/popv/.

Frequently Asked Questions about single-popv-annotation

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

FAQPage Schema
How do I get robust cell type annotation for scRNA-seq data when different algorithms give inconsistent results?▼

Robust cell type annotation for scRNA-seq data is achieved by running an ensemble of up to 10 classification algorithms, such as scVI and CellTypist, and aggregating their outputs through ontology-aware majority voting to produce consensus labels.

What is ontology-aware majority voting for scRNA-seq cell type annotation?▼

Ontology-aware majority voting for scRNA-seq cell type annotation aggregates multiple classification algorithm outputs by leveraging the Cell Ontology hierarchy to resolve labels and produce refined consensus predictions across reference atlases.

Can I use pretrained models for fast scRNA-seq cell type annotation on a new dataset?▼

Yes, you can use pretrained HubModel integration for fast scRNA-seq cell type annotation, enabling quick inference and atlas projection on new query datasets without needing to retrain the classification algorithms.

Do I need a GPU to run ensemble cell type annotation with scVI and XGBoost?▼

No, a GPU is not required to run ensemble cell type annotation, as the system supports CPU-only classical methods like RF and SVM for constrained environments, though optional GPU acceleration is available for faster inference.

Does the ensemble annotation pipeline handle automated preprocessing like normalization and HVG selection?▼

Yes, the ensemble annotation pipeline handles automated preprocessing including filtering, normalization, and HVG selection before executing classification algorithms to ensure query data matches the labeled reference format.

What is the best way to validate consensus cell-type labels generated from multiple annotation methods?▼

The best way to validate consensus cell-type labels is by generating agreement plots and prediction score summaries, which visualize the ensemble annotation agreement across algorithms to guide troubleshooting and method selection.

Related Skills

Explore More Data & Analytics Skills→
openyakopenyak
704

scvi-tools

Train scVI, scANVI, totalVI, PeakVI, and MultiVI models on single-cell data.

Official
Advanced
jaechang-hitsjaechang-hits
298

popv-cell-annotation

Annotate scRNA-seq cell types via consensus voting across ten algorithms.

Community
Advanced
StarlitnightlyStarlitnightly
32

single-popv-annotation

Annotate single-cell RNA sequencing data via multi-algorithm consensus voting.

Community
Advanced
Zeyuyang-0420Zeyuyang-0420

scvi-tools

Automate single-cell RNA-seq analysis with scvi-tools deep generative models.

Community
Advanced