About

Imported from lilinji/GeneTind-Life-Skills (skills/metabolomics-annotation/SKILL.md). Install upstream with npx skills add lilinji/GeneTind-Life-Skills --skill metabolomics-annotation. Copyright stays with the author (MIT).

metabolomics-annotation

Name: metabolomics-annotation
Author: lilinji

When to use

The user has a feature table with mz (m/z) values and wants each feature annotated by m/z match to a metabolite database. This is demo-only annotation. The reference is an 15-entry HMDB dictionary (metabolomics_annotation.py:57-74: Glucose, Lactic acid, Alanine, Glycine, Serine, Proline, Valine, Leucine). --database {hmdb,kegg,lipidmaps,metlin} is recorded as metadata but does NOT switch the lookup table.

For real database-scale annotation use SIRIUS / GNPS / MetFrag externally and feed the resulting annotation CSV into a downstream skill.

Inputs & Outputs

Inputs

File types: .csv

Outputs

tables/annotations.csv
report.md
result.json

Flow

Load CSV (--input <features.csv>) or generate a demo (--demo).
For each input mz, search the 15-entry HMDB dictionary (metabolomics_annotation.py:57-74) within --ppm tolerance.
Write tables/annotations.csv (metabolomics_annotation.py:279) + report.md + result.json.

Gotchas

Database is HARD-CODED 8 metabolites — --database is metadata only. metabolomics_annotation.py:57-74 defines an 15-entry HMDB tuple. The CLI accepts hmdb / kegg / lipidmaps / metlin (:251 choices=...) but the value is only logged into result.json — the lookup always uses the same 15-entry HMDB list. For real annotation, use SIRIUS / GNPS / MetFrag externally.
--ppm 10.0 default is m/z-tolerance. Suitable for high-resolution Orbitrap; for low-resolution Q-TOF use --ppm 30.0. The mass-error formula is |mz_obs - mz_ref| < (ppm × mz_ref / 1e6).
--input REQUIRED unless --demo. metabolomics_annotation.py:269 raises ValueError("--input required when not using --demo").
Required CSV column is mz (lowercase). XCMS exports mzmed, MZmine exports m/z; rename to mz first.
Multiple matches per feature ⇒ multiple rows. A feature with 3 candidate matches yields 3 rows in tables/annotations.csv; deduplicate downstream by feature_id if you need 1:1.

Key CLI

# Demo
python omicsclaw.py run metabolomics-annotation --demo --output /tmp/anno_demo

# Real feature table (annotates against demo HMDB dictionary regardless of --database)
python omicsclaw.py run metabolomics-annotation \
  --input features.csv --output results/ \
  --database hmdb --ppm 5.0

references/parameters.md — every CLI flag
references/methodology.md — m/z-match formula, demo-DB caveats
references/output_contract.md — tables/annotations.csv schema
Adjacent skills: metabolomics-xcms-preprocessing (upstream — feature × sample matrix), metabolomics-peak-detection (upstream — per-sample peak picking), metabolomics-quantification (parallel — impute + normalise), metabolomics-pathway-enrichment (downstream — pathway analysis on annotated features)

Use it

Copy one of these into your project. Installing also returns the manifest and these snippets.

yaml

targets:
  - https://api.opensmartroute.ai/api/v1/registry/lilinji-genetind-life-skills-metabolomics-annotation/manifest   # or paste the manifest below

Manifest

An Open Capability Manifest: the router reads it to know what this does, what it costs and when to pick it.

lilinji-genetind-life-skills-metabolomics-annotation.ocm.jsonjson

{
  "ocm": "1",
  "id": "lilinji-genetind-life-skills-metabolomics-annotation",
  "kind": "skill",
  "name": "metabolomics-annotation",
  "description": "Load when annotating LC-MS features against a built-in 15-metabolite HMDB demo dictionary by m/z within a `--ppm` tolerance — emits a per-feature annotation table. Skip when needing real HMDB / KEGG / LipidMaps / METLIN look-up (this skill is demo-only); raw spectra (use metabolomics-xcms-preprocessing).",
  "publisher": "lilinji",
  "version": "1.0.0",
  "capabilities": {
    "domains": [
      "data_analysis"
    ],
    "tags": [
      "skill-md",
      "metabolomics",
      "annotation",
      "hmdb",
      "demo",
      "mz-match",
      "github"
    ],
    "languages": [
      "en"
    ]
  },
  "quality_prior": 0.6,
  "examples": [
    "Load when annotating LC-MS features against a built-in 15-metabolite HMDB demo dictionary by m/z within a `--ppm` tolerance — emits a per-feature annotation table. Skip when needing real HMDB / KEGG / LipidMaps / METLIN look-up (this skill is demo-only); raw spectra (use metabolomics-xcms-preprocessing)."
  ],
  "primary": false,
  "metadata": {
    "source": {
      "provider": "github",
      "repository": "https://github.com/lilinji/GeneTind-Life-Skills",
      "path": "skills/metabolomics-annotation/SKILL.md",
      "ref": "908229c955510a1841b216e8bf58f6b22818d4dc",
      "url": "https://github.com/lilinji/GeneTind-Life-Skills/blob/908229c955510a1841b216e8bf58f6b22818d4dc/skills/metabolomics-annotation/SKILL.md",
      "key": "lilinji/GeneTind-Life-Skills/skills/metabolomics-annotation/SKILL.md"
    },
    "license": "MIT"
  },
  "instructions": "# metabolomics-annotation\n\n## When to use\n\nThe user has a feature table with `mz` (m/z) values and wants\neach feature annotated by m/z match to a metabolite database.\n**This is demo-only annotation.** The reference is an 15-entry\nHMDB dictionary (`metabolomics_annotation.py:57-74`: Glucose,\nLactic acid, Alanine, Glycine, Serine, Proline, Valine, Leucine).\n`--database {hmdb,kegg,lipidmaps,metlin}` is recorded as metadata\nbut does NOT switch the lookup table.\n\nFor real database-scale annotation use SIRIUS / GNPS / MetFrag\nexternally and feed the resulting annotation CSV into a downstream\nskill.\n",
  "cost": {
    "context_tokens": 755
  }
}

Fetch it by URL: GET /api/v1/registry/lilinji-genetind-life-skills-metabolomics-annotation/manifest?version=1.0.0

Reviews

Star ratings from people who tried it. One review per account; edit yours any time.

No reviews yet. Install it, try it, and be the first to rate it.