# Metabolite Pathway Association Mapping

> Use when you have a list of metabolite names or identifiers detected in your samples and need to assign them to known metabolic pathways before computing pathway dysregulation scores, performing pathway-level machine learning, or conducting metabolite-pathway regression analysis.

- Skill: `holobiomicslab/metabolite-pathway-association-mapping` (Agent Skill)
- Install (CLI): `npx skillmds@latest add holobiomicslab/metabolite-pathway-association-mapping`
- Raw SKILL.md: https://api.skillmd.com/api/skills/holobiomicslab/metabolite-pathway-association-mapping/raw
- Safety review: PASS (external: skill-scanner PASS, skillspector PASS)
- Works with: Claude Code, Claude.ai, OpenAI Codex
- Category: Coding & Dev Tools
- License: CC-BY-4.0
- Author: HolobiomicsLab (https://skillmd.com/u/holobiomicslab)
- Updated: 2026-09-17
- Page: https://skillmd.com/skills/holobiomicslab/metabolite-pathway-association-mapping

---


# metabolite-pathway-association-mapping

## Summary

Transform individual metabolite identifiers (names or masses) into a standardized metabolite-pathway association table that maps each metabolite to one or more canonical metabolic pathways. This is a prerequisite step for pathway-level analysis in personalized metabolomics studies.

## When to use

You have a list of metabolite names or identifiers detected in your samples and need to assign them to known metabolic pathways before computing pathway dysregulation scores, performing pathway-level machine learning, or conducting metabolite-pathway regression analysis.

## When NOT to use

- Your metabolites are already pre-mapped to pathways in a curated database and require no conversion.
- You are working with a non-standard organism or custom metabolite nomenclature not covered by Lilikoi's embedded or linked databases.
- Your analysis goal is sample-level classification or prognosis without pathway-level aggregation; metabolite-level features may suffice.

## Inputs

- List or vector of metabolite names (string identifiers)
- Sample metabolite abundance data (optional, for validation context)

## Outputs

- Metabolite-pathway association table (rows: metabolites; columns: pathway assignments)
- Standardized metabolite identifiers (e.g., HMDB IDs)

## How to apply

Use Lilikoi's lilikoi.MetaTOpathway function to convert metabolite names to standardized identifiers (such as HMDB IDs) and retrieve their pathway assignments from a curated metabolite-pathway database. The function outputs a table with metabolite identifiers as rows and pathway annotations as columns (or a lookup structure). This association table becomes the reference used downstream by lilikoi.PDSfun (for dysregulation scoring), lilikoi.meta_path (for regression), and lilikoi.KEGGplot (for visualization). The conversion ensures consistency across the analysis pipeline and enables aggregation of metabolite-level measurements into pathway-level features.

## Related tools

- **Lilikoi v2.0** (R package providing lilikoi.MetaTOpathway function for metabolite name-to-pathway conversion and lilikoi.PDSfun for downstream pathway dysregulation scoring) — https://github.com/lanagarmire/lilikoi2
- **R** (Programming environment in which Lilikoi v2.0 executes the metabolite-pathway mapping workflow)
- **HMDB (Human Metabolome Database)** (Referenced database providing standardized metabolite identifiers and pathway annotations used by lilikoi.MetaTOpathway)

## Examples

```
convertResults=lilikoi.MetaTOpathway('name'); Metabolite_pathway_table = convertResults$table; head(Metabolite_pathway_table)
```

## Evaluation signals

- Output table has no null or unmapped metabolites; all input metabolite names resolve to standardized identifiers (e.g., HMDB IDs).
- Pathway annotations are non-empty for >95% of metabolites; the table is sparse but complete (no column of all zeros).
- Downstream lilikoi.PDSfun executes without errors, confirming that the table structure and metabolite identifiers are compatible.
- Spot-check: manually verify that 2–3 known metabolites (e.g., glucose, pyruvate, lactate) are assigned to expected pathways (e.g., Glycolysis, TCA cycle).
- Table dimensions are consistent with input metabolite count and the expected size of the organism's metabolic network (e.g., hundreds of metabolites × dozens of pathways for human data).

## Limitations

- Lilikoi's metabolite name conversion is limited to metabolites in curated pathway databases (primarily human metabolism); rare, novel, or organism-specific metabolites may not be mapped.
- Metabolite name ambiguity (e.g., common chemical names vs. IUPAC nomenclature) can lead to incorrect or missed mappings; standardized nomenclature (e.g., HMDB IDs) is recommended as input.
- The association table is static at the time of analysis; updates to pathway annotations or database content do not propagate automatically.

## Evidence

- [readme] Transform the metabolite names to the HMDB ids using Lilikoi MetaTOpathway function: "Transform the metabolite names to the HMDB ids using Lilikoi MetaTOpathway function"
- [other] For each pathway, extract the subset of metabolites assigned to that pathway.: "For each pathway, extract the subset of metabolites assigned to that pathway."
- [intro] Lilikoi v2 supports data preprocessing, exploratory analysis, pathway visualization and metabolite-pathway regression: "Lilikoi v2 supports data preprocessing, exploratory analysis, pathway visualization and metabolite-pathway regression."
- [other] Organize scores into a matrix with samples as rows and pathways as columns, generating the PDSmatrix output.: "Organize scores into a matrix with samples as rows and pathways as columns"