CompoundDb

CompoundDb creates and uses chemical compound annotation databases from sources such as LipidMaps, HMDB, ChEBI, or MassBank, storing MS/MS spectra alongside compound information and providing a backend for Bioconductor's Spectra package to match experimental MS/MS spectra against database spectra.

Key Features:

• Database Creation: CompoundDb generates compound annotation databases from diverse sources including LipidMaps, HMDB, ChEBI, and MassBank.
• MS/MS Spectra Storage: The database format stores tandem mass spectrometry spectra together with compound information.
• Spectra Package Backend: CompoundDb serves as a backend for Bioconductor's Spectra package, enabling matching of experimental MS/MS spectra against reference database spectra.
• Portable Database Format: Databases are stored in SQLite format for portability.
• Mass Matching Annotation: CompoundDb performs tentative annotations based on mass matching for metabolomics datasets.
• LC-MS Feature Annotation: The package considers m/z and retention times for annotation of LC-MS features when reference values are available.
• LC-MS/MS Spectral Library Matching: CompoundDb provides high-level functions to match LC-MS/MS spectra against spectral libraries and to represent and manage matched data.
• Multiple Data Format Support: The package supports data formats including MSP, MGF, mzML, mzXML, netCDF, MassBank text files, and SQL databases.

Scientific Applications:

• Untargeted Metabolomics Annotation: CompoundDb enables modular infrastructure for annotation of untargeted metabolomics data through MS1-based annotation using m/z and retention times followed by MS2-based annotation comparing experimental fragment spectra against reference libraries.
• Metabolomics Data Set Annotation: The package assists in annotation of metabolomics datasets through mass matching and retention time-based approaches.
• Reproducible Annotation Workflows: CompoundDb supports construction of customizable, reproducible annotation workflows adapted to diverse untargeted LC-MS-based datasets.

Methodology:

CompoundDb creates annotation databases by importing compound information from external sources, stores these in SQLite format with associated MS/MS spectra, and matches experimental MS/MS spectra against database reference spectra through integration with Bioconductor's Spectra package. Annotation proceeds from initial MS1 matching based on m/z and retention time to subsequent MS2-level annotation through comparison of experimental fragment spectra against reference spectral libraries.