agentsclimarketplace

Multi database structure querying

Skill HolobiomicsLab/asb-skill-collections/packs/metabolomics/lc-ms/skills/multi-database-structure-querying

Curated, evidence-grounded skill and software-tool collections for scientific AI agents, generated by the AgenticScienceBuilder

Install
npx -y skills add HolobiomicsLab/asb-skill-collections --skill multi-database-structure-querying

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

  • 14 stars14 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 author says it does

Copied from the file, not written here

Use when after cleaning and clustering LC-MS features in MS-CleanR, when you need to assign putative compound identities to a feature set with MS/MS spectra, and when candidate compounds may exist across multiple specialized databases (e.g., natural products, pharmaceuticals, contaminants).

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

8.3 KB, ~1.5k tokens by cl100k_base, as published. Nobody here has run it

multi-database-structure-querying

Summary

Query multiple metabolite structure databases simultaneously via MS-FINDER to annotate LC-MS features using hydrogen rearrangement rules (HRR) scoring. This skill enables parallel, database-agnostic compound identification from cleaned and clustered MS-DIAL peak lists, with consolidation of per-feature annotations into a unified table ranked by user-defined database priority.

When to use

After cleaning and clustering LC-MS features in MS-CleanR, when you need to assign putative compound identities to a feature set with MS/MS spectra, and when candidate compounds may exist across multiple specialized databases (e.g., natural products, pharmaceuticals, contaminants). Use this skill when you want to leverage HRR-based scoring to rank structural candidates and compare results across databases rather than querying a single database sequentially.

When NOT to use

  • Input features lack MS/MS spectra — MS-CleanR will discard all features without MS/MS during the first filtering step, and MS-FINDER requires fragmentation data for in silico annotation.
  • Data were acquired in MS1-only mode (no tandem MS) — the HRR scoring system and database querying depend on fragmentation patterns.
  • You have already committed to a single reference database and do not need candidate ranking or cross-database comparison.

Inputs

  • MS-CleanR-filtered and clustered LC-MS feature set with m/z, retention time, and MS/MS spectra
  • MS-FINDER–compatible input format (feature table with spectral data)
  • Configuration file or parameters specifying target databases and HRR scoring settings

Outputs

  • Unified annotation table per feature (compound ID, HRR score, database source, structural predictions)
  • Merged feature table with annotations prioritized by user-defined database rank
  • Optional .msp file for mass spectral networking

How to apply

Export the selected feature set (m/z, retention time, MS/MS spectra) from MS-CleanR in MS-FINDER–compatible format. Configure MS-FINDER annotation parameters to enable the hydrogen rearrangement rules (HRR) scoring system and specify multiple target databases for querying in a single batch run. Execute MS-FINDER in silico annotation on the exported features. Parse and consolidate all per-feature annotation results (compound ID, HRR score, database match, structural predictions) into a unified annotation table. Merge annotation results back to the filtered peak list by prioritizing database hits according to user-defined ranking criteria (e.g., highest HRR score, preferred database order). Optionally export consolidated results as .msp file for mass spectral similarity networking.

Related tools

Evaluation signals

  • Every feature in the input set receives at least one HRR-scored annotation with a database source label; no features are silently dropped.
  • Per-feature annotation records include compound ID, HRR score, database source, and structural predictions; no field is missing or null for matched features.
  • Database prioritization ranking is consistently applied across all features (e.g., a user-specified database always ranks first when present, or highest HRR score always wins ties).
  • Merged feature table row count matches input feature count (or documents any intentional exclusions, e.g., unmatched features).
  • Optional .msp export file parses without syntax errors and contains spectral entries matching the consolidated annotation count.

Limitations

  • Requires at least MS-DIAL v4.00 and MS-FINDER v3.30 or higher; newer MS-DIAL 5.x integration may supersede this workflow as MS-CleanR is no longer actively maintained.
  • All features without MS/MS spectra are discarded in the first MS-CleanR filtering step; MS1-only data will cause a crash.
  • At least 3 blank and 3 QC samples must be identified in the MS-DIAL sample list for proper blank ratio analysis during feature cleaning.
  • Sample and class names must not contain spaces, hyphens, or single-letter class identifiers, or the workflow will fail.
  • Known bug in R > 4.2: 'Error: the condition has length > 1' encountered during database annotation merging; workaround or R version downgrade required.
  • HRR scoring provides in silico predictions; annotations are not validated against experimental standards and should be confirmed by orthogonal methods (e.g., high-resolution MS, authentic standards, NMR).

Evidence

  • [readme] all selected features are exported to MS-FINDER program for in silico-based annotation using hydrogen rearrangement rules (HRR) scoring system. At this step, multiple databases can be queried: "all selected features are exported to MS-FINDER program for in silico-based annotation using hydrogen rearrangement rules (HRR) scoring system. At this step, multiple databases can be queried"
  • [readme] The final step will merge annotation results to the filtered peak list by prioritizing database annotation depending on user choice.: "The final step will merge annotation results to the filtered peak list by prioritizing database annotation depending on user choice."
  • [other] MS-CleanR exports selected features to MS-FINDER for in silico-based annotation using hydrogen rearrangement rules (HRR) scoring system, with capability to query multiple databases at this step.: "MS-CleanR exports selected features to MS-FINDER for in silico-based annotation using hydrogen rearrangement rules (HRR) scoring system, with capability to query multiple databases at this step."
  • [other] Parse MS-FINDER results and consolidate per-feature annotations (compound ID, score, database match, structural predictions) into a unified annotation table.: "Parse MS-FINDER results and consolidate per-feature annotations (compound ID, score, database match, structural predictions) into a unified annotation table."
  • [readme] All these options are tunable by the user.: "All these options are tunable by the user."
  • [readme] All features without MS/MS will be discarded during the first step. If data contain MS1 only, the first MS-CleanR step will crash.: "All features without MS/MS will be discarded during the first step. If data contain MS1 only, the first MS-CleanR step will crash."
  • [readme] Optionally, all results can be exported as .msp file for mass spectral similarity networking purpose.: "Optionally, all results can be exported as .msp file for mass spectral similarity networking purpose."

Keep looking

Skills are one crate of 328,083. 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.