agentsclimarketplace

Classyfire taxonomy assignment

Skill HolobiomicsLab/asb-skill-collections/collections/metabolomics/v2/skills/classyfire-taxonomy-assignment

Use when after molecular structures have been standardized (e.g., via PubChem standardization) and you need to enrich them with chemical taxonomy labels for cohort stratification, chemical space analysis, or retention time prediction model development.From its SKILL.md

Install
npx -y skills add HolobiomicsLab/asb-skill-collections --skill classyfire-taxonomy-assignment

Assembled from the repository path, not quoted from the project. Check it against their README if it does not work.

One thing to look at

  • 15 stars15 stars. Stars are a popularity signal and not a quality one, but at this level it is likely that nobody has read this closely except its author, and you would be relying on your own review.

What its file declares

Copied from the file, not written here

The file declares its own license as CC-BY-4.0. That is the author’s claim about this one file, and it is not the same thing as the license GitHub reports for the repository, which is listed with the other numbers below.

SKILL.md

6.2 KB, 892 tokens by cl100k_base, as published. Nobody here has run it

classyfire-taxonomy-assignment

Summary

Assign standardized chemical taxonomy (kingdom, superclass, class, subclass) to molecular structures using the ClassyFire API or local instance. This skill is essential in cheminformatics pipelines to enable downstream chemical classification-based analysis and filtering of small molecules.

When to use

After molecular structures have been standardized (e.g., via PubChem standardization) and you need to enrich them with chemical taxonomy labels for cohort stratification, chemical space analysis, or retention time prediction model development. Apply when your input is a collection of standardized molecular structures (SMILES or MOL format) lacking kingdom/superclass/class/subclass annotations.

When NOT to use

  • Input structures have not been standardized (use PubChem standardization first)
  • Molecules are already pre-classified with an alternative taxonomy system and no re-annotation is desired
  • Internet or API access to ClassyFire is unavailable and no local instance is set up

Inputs

  • Standardized molecular structures (SMILES or MOL format)
  • Molecular identifiers or batch file of structures
  • Connection parameters to ClassyFire API or local instance

Outputs

  • Aggregated classification table (CSV or JSON) with molecular identifiers and ClassyFire taxonomy (kingdom, superclass, class, subclass)
  • Validation report flagging molecules with null or incomplete classification records

How to apply

Load standardized molecular structures (SMILES or MOL format) from input file. Query the ClassyFire API or local instance with each structure to retrieve the complete chemical classification taxonomy. Aggregate the returned classification results (kingdom, superclass, class, subclass) with corresponding molecular identifiers into a structured output table (CSV or JSON). Validate that each molecule has a non-null ClassyFire classification record and all available taxonomy levels are populated; flag or retry any molecules that fail classification or return incomplete taxonomy branches.

Related tools

  • ClassyFire (Queries and retrieves standardized chemical classification taxonomy for each molecular structure)
  • PubChem standardization (Pre-processes and standardizes input molecular structures before ClassyFire annotation)
  • rcdk (Calculates molecular fingerprints and chemical descriptors, often used in conjunction with ClassyFire taxonomy for comprehensive molecular characterization)

Evaluation signals

  • All input molecules have a corresponding non-null ClassyFire classification record in the output table
  • Taxonomy levels (kingdom, superclass, class, subclass) are populated where available; no incomplete or truncated taxonomy branches
  • Aggregated output table structure matches specified format (CSV or JSON) with all expected columns (molecular ID, kingdom, superclass, class, subclass)
  • Validation report contains zero 'failed classification' flags; any molecules flagged should be manually inspected or retried
  • Spot-check: verify that known reference structures (e.g., a lipid, amino acid, or drug compound) are assigned to chemically plausible taxonomy nodes

Limitations

  • ClassyFire classification relies on external API availability or a maintained local instance; API rate limits or downtime may delay large batch processing
  • Non-standard or highly unusual molecular structures may fail to classify or receive incomplete taxonomy annotations; such molecules should be manually reviewed
  • Classification accuracy depends on the quality of input structure standardization; errors in structure standardization propagate to taxonomy assignment

Evidence

  • [readme] structures are standardized using the PubChem standardization and molecular fingerprints and chemical descriptors are calculated using rcdk. Classification of molecules is performed using ClassyFire.: "structures are standardized using the PubChem standardization and molecular fingerprints and chemical descriptors are calculated using rcdk. Classification of molecules is performed using ClassyFire."
  • [other] Query ClassyFire API or local instance with each structure to retrieve chemical classification taxonomy (kingdom, superclass, class, subclass).: "Query ClassyFire API or local instance with each structure to retrieve chemical classification taxonomy (kingdom, superclass, class, subclass)."
  • [other] Aggregate classification results with corresponding molecular identifiers into a structured table (CSV or JSON).: "Aggregate classification results with corresponding molecular identifiers into a structured table (CSV or JSON)."
  • [other] Validation: verify each molecule has a non-null ClassyFire classification record and all taxonomy levels are populated where available.: "Validation: verify each molecule has a non-null ClassyFire classification record and all taxonomy levels are populated where available."

What ships with it

Read from the repository

Just SKILL.md. No reference files, no scripts.

Keep looking

Skills are one crate of 325,949. Ordering is by how many stacks a row turns up in, so the top of any crate is what has actually been picked rather than what has the most stars.