Untargeted LC-MS/MS metabolomics can detect thousands of molecular features across complex biological samples. However, turning those signals into biologically meaningful metabolite annotations requires more than accurate-mass measurement alone.

This technical note explores how database evidence, MS/MS fragmentation and confidence-ranked reporting can support more interpretable untargeted metabolomics results. It introduces NovoMetDB-UM-V4.0, Novogene’s in-house metabolomics database, and explains how reference-standard data, curated public spectra and AI-predicted spectra contribute to metabolite annotation.

The note also outlines the distinction between confirmed Level 1 identifications, probable Level 2 annotations and exploratory feature-level signals, alongside the quality-control processes used to assess analytical stability and reproducibility.

What you will learn:

• How LC-MS/MS features are processed for metabolite annotation
• How NovoMetDB-UM-V4.0 supports spectral matching
• What different metabolite annotation confidence levels mean
• How pooled QC samples, blanks and internal standards support data quality
• What researchers receive from an untargeted metabolomics project