Metagenomics Skills

简体中文

Portable Agent Skills for reproducible shotgun metagenomics, organized by
scientific stage and usable from Claude Code, Codex, Cursor, Kimi Code, ZCode,
Qwen Code, and other compatible coding agents:

QC → assembly → mapping → binning → taxonomy → function → microbial mining → research analysis

One skill per core tool (mappers, assemblers, binners, profilers, miners such as
antiSMASH / geNomad / DRAM, …) plus databases, language-model hubs, homology /
phylogenomics routing, and paper-facing stats/figures (research-analysis).

Getting started ·
Skills ·
Pipeline ·
Public databases ·
Agent playbook ·
References ·
Platforms

Install

Claude Code

/plugin marketplace add PKU-EMBL/Metagenomics-Skills
/plugin install metagenomics@metagenomics-skills

Codex plugin

codex plugin marketplace add PKU-EMBL/Metagenomics-Skills
codex plugin add metagenomics@metagenomics-skills

Directory install: Codex / Cursor / Kimi / ZCode / Qwen / Grok / custom agents

git clone https://github.com/PKU-EMBL/Metagenomics-Skills.git
cd Metagenomics-Skills

# Recommended: compact high-use catalog for Kimi Code, ZCode, and Qwen Code
./scripts/install-platforms.sh --user --profile core kimi zcode qwen

# Shared path for Codex + Kimi + ZCode, plus Qwen native path
./scripts/install-platforms.sh --user --profile core agents qwen

# Or install every skill for every known platform
./scripts/install-platforms.sh --user --profile full all

Use --target PATH for a new client, --mode copy when symlinks are not
supported, --dry-run to preview, and --prune when switching an existing
destination to a smaller profile. See
the complete platform matrix.

Layout

skills/<tool>/SKILL.md       # canonical portable skill
skills/<tool>/references/    # optional progressive detail, kept in-bundle
config/skill-profiles.json   # low-context and task-focused install sets
docs/                        # catalogs, pipeline, agent execution contract
plugin.json                  # portable Agent Plugins v1 package
qwen-extension.json          # Qwen Code extension
.claude-plugin/ .codex-plugin/ .cursor-plugin/ .zcode-plugin/
scripts/install-skills.py    # safe link/copy/custom-target installer
scripts/validate-skills.py   # spec/manifests/profiles/links validator

Quick start

Tell the agent whether to advise, plan, prepare, execute, or
review. It should inspect inputs, preflight tools/databases/resources, pilot,
validate every stage boundary, and report what was actually run. See the
coding-agent playbook.

Full genome-resolved path

study-design → QC (± multiqc) → assembly → mapping →
binning (basalt recommended) → (± anvio / deepurify) →
taxonomy / function → research-analysis (microeco / maaslin2 / …)

Community composition only (no assembly)

QC → taxonomy (metaphlan, singlem, or kraken2) →
research-analysis

Research stats & figures

research-analysis → multiqc / microeco / phyloseq / visualization /
maaslin2 / ancombc / lefse / stamp / sparcc / spieceasi

Workflow managers

nextflow (± nf-core) · snakemake

Genome-resolved + interactive polish

Auto-bins → anvio refine → checkm2 / gunc / mag-qc → gtdbtk

Viral / plasmid / MGE mining

Contigs → genomad / virsorter2 → checkv → vcontact ·
mobileog / isescan / integronfinder / mob-suite · plasmids: plasmaag ·
isolates: cge-finders (staramr / abricate / plasmidfinder → pmlst / mlst)

Microbial mining (BGCs / AMR / element cycling / DOM / defense / QS)

After MAGs: microbial-mining → antismash / rgi / cge-finders /
element-cycling / dom-biosurfactant / dram / defensefinder / padloc /
qsap / …

Homology search (Steinegger/Söding · vector DB · compressed genomes)

homology-search → mmseqs2 / hh-suite / foldseek / foldmason /
plass / petasearch / erast / phylign / …

Phylogenomics / trees

phylogenomics → gtdbtk / phylophlan / unicore / iqtree / fasttree /
lucaphylo

Genome language models

genome-language-model → omg / nucleotide-transformer (NTv3) /
dnabert2 / caduceus / evo2

Protein language models

protein-language-model → esm / esmc / saprot / prostt5 /
plmsearch / deepblast

AI / LMs & structure search (Foldseek · PLMs)

metagenomics-llm → genome-language-model / protein-language-model /
foldseek / spacedust / …

Contributing

See CONTRIBUTING.md. New skills go under skills/<name>/
with upstream links and an “Analytical thinking” section. Validate changes with
python3 scripts/validate-skills.py.

License

MIT

References

Primary literature (Nature-style). Full numbered list:
docs/references.md.

1. Qiu, Z. et al. BASALT refines binning from metagenomic data and increases resolution of genome-resolved metagenomic analysis. Nat. Commun. 15, 2179 (2024). https://doi.org/10.1038/s41467-024-46539-7

2. Uritskiy, G. V., DiRuggiero, J. & Taylor, J. MetaWRAP—a flexible pipeline for genome-resolved metagenomic data analysis. Microbiome 6, 158 (2018). https://doi.org/10.1186/s40168-018-0541-1

3. Sieber, C. M. K. et al. Recovery of genomes from metagenomes via a dereplication, aggregation and scoring strategy. Nat. Microbiol. 3, 836–843 (2018). https://doi.org/10.1038/s41564-018-0171-1

4. Blanco-Míguez, A. et al. Extending and improving metagenomic taxonomic profiling with uncharacterized species using MetaPhlAn 4. Nat. Biotechnol. 41, 1633–1644 (2023). https://doi.org/10.1038/s41587-023-01688-w

5. Beghini, F. et al. Integrating taxonomic, functional, and strain-level profiling of diverse microbial communities with bioBakery 3. eLife 10, e65088 (2021). https://doi.org/10.7554/eLife.65088

6. Almeida, A. et al. A unified catalog of 204,938 reference genomes from the human gut microbiome. Nat. Biotechnol. 39, 105–114 (2021). https://doi.org/10.1038/s41587-020-0603-3

7. Parks, D. H. et al. A complete domain-to-species taxonomy for Bacteria and Archaea. Nat. Biotechnol. 38, 1079–1086 (2020). https://doi.org/10.1038/s41587-020-0501-8

8. Richardson, L. et al. MGnify: the microbiome sequence data analysis resource in 2023. Nucleic Acids Res. 51, D753–D759 (2023). https://doi.org/10.1093/nar/gkac1080

Software repositories and Agent Skills packaging links are listed in
docs/references.md only as implementation pointers; cite the
papers above (plus versions/releases) in manuscripts.