HolobiomicsLab cross-domain-metadata-integration
Summary
Consolidates and visualizes aggregated mass spectrometry search results from multiple domain-specific MASST tools (microbeMASST, plantMASST, tissueMASST, microbiomeMASST, foodMASST) into unified metadata structures and interactive visualizations. This skill enables cross-domain interpretation of molecular matches and taxonomic/ecological overlap across curated biological repositories.
When to use
When you have conducted batch MS/MS searches across one or more domain-specific MASST tools and need to combine their hit scores, metadata annotations, and taxonomic lineages into a single coherent result set for comparative analysis or publication. Specifically: after generating domain-specific .json/.html tree outputs from jobs.py, or when merging results from multiple domain repositories (GNPS, MassIVE, Metabolomics Workbench, Metabolights, NORMAN) and require unified visualization of cross-domain matches, sample overlap, and scoring distributions.
When NOT to use
- Input consists of only single-domain results (i.e., only microbeMASST or only plantMASST); cross-domain integration provides no additional signal.
- You seek domain-specific curation or taxonomic filtering that metadataMASST does not expose; use individual domain-MASST web applications instead.
- Raw MS/MS spectra have not yet been searched; first run batch search via jobs.py against Fast Search API.
Inputs
- aggregated _matches.tsv (all spectra matching results across indexed databases)
- aggregated _library.tsv (GNPS library matches for Level 2 MSI annotation)
- aggregated _datasets.tsv (unique sample counts per indexed dataset)
- per-domain _count_domain.tsv files (microbe, plant, tissue, microbiome, food)
- batch search output JSON trees from individual domainMASSTs
Outputs
- consolidated metadata table with merged hits, scores, and lineage annotations
- interactive HTML visualization(s) of aggregated result (domain overlap, hit distribution)
- JSON structure encoding cross-domain match graph and metadata hierarchy
- domain-wise summary report (e.g., match counts, score distributions per domain)
How to apply
Load the aggregated search outputs (typically _matches.tsv, _library.tsv, _datasets.tsv, and per-domain _count_domain.tsv files) from the batch search into metadataMASST. The tool processes and consolidates hits by merging cosine scores, m/z matches, and metadata annotations across the different domain sources. Adjust integration parameters such as minimum cosine score threshold (typically 0.6–0.7 for MS/MS) and m/z tolerance (default 0.5 Da) based on your mass spectrometry instrument resolution and biological specificity goals. Generate visualization outputs (e.g., hit distribution histograms, domain-overlap Venn diagrams, score ranking tables) and export in structured formats (interactive HTML trees, JSON for downstream processing). The rationale is that domain-specific MASSTs each curate different NCBI taxonomy lineages and sample metadata; integration reveals which molecular features are present across multiple biological domains, improving confidence in annotations and enabling ecosystem-scale metabolomics interpretation.
Related tools
- metadataMASST (primary integration and visualization engine; consolidates and visualizes aggregated cross-domain search results) — https://masst.gnps2.org/metadatamasst/
- microbeMASST (domain-specific MASST for microbial metabolomics; produces indexed hits and taxonomic lineages for bacteria/archaea integration) — https://masst.gnps2.org/microbemasst/
- plantMASST (domain-specific MASST for plant metabolomics; produces indexed hits and plant taxonomic lineages for integration) — https://masst.gnps2.org/plantmasst/
- tissueMASST (domain-specific MASST for tissue and organ metabolomics; produces hits and tissue metadata for cross-domain matching) — https://masst.gnps2.org/tissuemasst/
- microbiomeMASST (domain-specific MASST for microbiome metabolomics; produces ecological and taxonomic metadata for integration) — https://masst.gnps2.org/microbiomemasst/
- foodMASST (domain-specific MASST for food metabolomics; produces food source and provenance metadata for cross-domain matching) — https://masst.gnps2.org/foodmasst2/
- jobs.py (batch search pipeline that generates aggregated domain MASST outputs (_matches.tsv, _library.tsv, _datasets.tsv, per-domain _count_domain.tsv) for integration) — https://github.com/robinschmid/microbe_masst/blob/master/code/jobs.py
- Fast Search API (underlying search infrastructure leveraged by jobs.py to execute batch MS/MS spectrum matching against GNPS, MassIVE, Metabolomics Workbench, Metabolights, and NORMAN databases) — https://fasst.gnps2.org/fastsearch/
- GNPS_MASST (parent repository containing code and logic for domain-specific MASST web applications and metadataMASST integration workflows) — https://github.com/mwang87/GNPS_MASST
Examples
python jobs.py --input_file batch_spectra.mgf --output_prefix results/cross_domain --min_cosine_score 0.6 --mz_tolerance 0.5 --skip_existing=True
Evaluation signals
- Verify that _matches.tsv, _library.tsv, and all _count_domain.tsv files are successfully loaded and deduplicated (no duplicate spectra IDs across domains).
- Check that cosine scores are correctly merged and ranked: scores from the same spectrum matched across multiple domains should show consistent ordering within metadataMASST output.
- Confirm that domain-overlap visualizations (e.g., HTML tree or Venn diagram) show non-empty intersection sets when multiple domains report hits for the same query spectrum.
- Validate that lineage information (kingdom, phylum, class, order, family, genus, species) is preserved and correctly attributed to matched spectra in consolidated metadata.
- Inspect exported JSON structure for completeness: all input domain source labels, metadata fields, and score tuples should be present and queryable.
Limitations
- metadataMASST assumes pre-indexed data across domain-specific MASSTs; newly submitted spectra or datasets require re-running the batch search pipeline (jobs.py) and multiple re-runs due to Fast Search API transient failures.
- Domain lineage coverage is uneven: microbeMASST currently covers ~1,379 microbial species and 542 strains; plantMASST covers ~3,712 plant species but no strains; tissue and microbiome/food MASST lineage depths are not quantified in the article.
- Cross-domain integration does not perform de novo curation or conflict resolution: if the same spectrum is annotated with conflicting taxonomic assignments across domains, metadataMASST surfaces all annotations without algorithmic prioritization.
- Batch search performance and completeness depend on Fast Search API availability; retry logic (skip_existing=True) is recommended but not guaranteed to catch all matches in a single execution.
Evidence
- [other] metadataMASST takes aggregated search outputs from domain-specific MASSTs... and generates visualized aggregated results: "metadataMASST takes aggregated search outputs from domain-specific MASSTs (microbeMASST, plantMASST, tissueMASST, microbiomeMASST, and foodMASST) and generates visualized aggregated results as its"
- [other] Process and consolidate search results using metadataMASST to merge hits, scores, and metadata across domain sources: "Process and consolidate search results using metadataMASST to merge hits, scores, and metadata across domain sources."
- [readme] Aggregated search outputs can be generated and visualized using metadataMASST: "Aggregated search outputs can be generated and visualized using metadataMASST."
- [readme] Running jobs.py allows users to leverage the Fast Search API and execute a batch search of multiple MS/MS spectra against the current indexed data in GNPS/MassIVE, Metabolomics Workbench, Metabolights, and NORMAN and generate multiple outputs for all listed domainMASSTs simultaneously: "Running jobs.py allows users to leverage the Fast Search API and execute a batch search of multiple MS/MS spectra against the current indexed data in GNPS/MassIVE, Metabolomics Workbench,"
- [readme] A _matches.tsv file will be generated. This contains all the scans found to match your searched spectrum of interest: "A _matches.tsv file will be generated. This contains all the scans found to match your searched spectrum of interest in the data that have been currently indexed."
- [readme] Make sure to run jobs.py a couple of times, until no new output is generated by having the option:
skip_existing=True. Due to the Fast Search API some of the entries will fail.: "Make sure to run jobs.py a couple of times, until no new output is generated by having the option:skip_existing=True. Due to the Fast Search API some of the entries will fail."