MS-Net: Annotated Metabolite Features from Cannabis Sativa Extracts

1275 metabolite annotations for 1297 features derived from Cannabis sativa extracts using a multi-similarity network workflow. The MS-Net workflow, developed by Victor Pereira Francisco, integrates mass spectral similarity, molecular structure, and taxonomic knowledge to prioritize annotations from a candidate space of over 118,000 structures. It successfully reconstructed known cannabinoid biosynthetic pathways, validating its biological coherence.

Use Cases

• Benchmarking metabolite annotation algorithms based on the multi-similarity network approach described.
• Training machine learning models for compound identification based on integrated spectral and structural similarity metrics.
• Validating biological pathway reconstruction in plant metabolomics based on the described cannabinoid biosynthetic pathway results.
• Prioritizing candidate structures in untargeted metabolomics studies based on the composite Link Score methodology.

Strengths

• 1275 high-confidence compound annotations assigned from an initial candidate space of over 118,000 structures.
• 53% of final annotations were rescued from lower-ranked positions (ranks 2–50), demonstrating correction of initial in silico ranking.
• Workflow validation includes successful reconstruction of known cannabinoid biosynthetic pathways.

Limitations

• Column-level documentation is absent; field semantics must be inferred after download.
• Row count is unknown, which may limit suitability assessment for specific modeling tasks.
• The dataset's scope is limited to metabolite features from Cannabis sativa extracts.

Provenance

Source

figshare

Collection Method

Generated via the MS-Net computational workflow applied to untargeted LC-MS data from Cannabis sativa extracts.

Freshness

Last updated 2026-05-30 13:04:42; freshness should be verified.