Lc Ms Dataset Acquisition And Curation

LC-MS Dataset Acquisition and Curation

Summary

Systematic retrieval, preparation, and organization of centroid mzML LC-MS metabolomics datasets from public repositories (MetaboLights, MassIVE) or vendor instruments, with format conversion and sample registry construction to enable downstream feature detection and scalability benchmarking.

When to use

When beginning an untargeted LC-MS metabolomics study and need to assemble a cohort of mzML files for processing; particularly when establishing performance baselines across sample counts (10, 50, 100+ samples), validating reproducibility, or preparing data for publication. Also apply when converting vendor-native formats (Thermo .RAW, Waters, Bruker) to open mzML format.

When NOT to use

• If input data are already in profile mode (not centroid) and centroid peak picking has not been performed — asari requires centroid mzML
• If MS/MS spectra are the primary analytical target and no MS1 feature detection is needed — consider specialized MS/MS workflows instead
• If sample cohort size is <10 and scalability benchmarking is the goal — minimum sample counts of 10, 50, 100+ are recommended for valid runtime/memory trend analysis

Inputs

• Public mzML datasets from MetaboLights or MassIVE repositories
• Vendor raw data files (Thermo .RAW, Waters .raw, Bruker .d)
• Sample metadata (IDs, ionization mode, acquisition settings)

Outputs

• Organized project directory with centroid mzML files
• Sample registry with metadata and file paths
• Indexed mzML files (if conversion was performed)
• Documentation of MS1 scan structure and m/z/RT ranges

How to apply

Retrieve centroid mzML datasets from MetaboLights or MassIVE public repositories, or acquire vendor raw files from your instrument. Convert non-mzML vendor formats (e.g. Thermo .RAW) to indexed mzML using ThermoRawFileParser or ProteoWizard msconvert. Organize files in a project directory structure. Build a sample registry by recording metadata (sample ID, instrument mode, acquisition parameters) in a structured format. Verify all files are centroid-mode (not profile) mzML with complete MS1 spectra intact; asari ignores MS/MS spectra in default LC-MS workflow. Document MS1 scan structure, retention time range, and m/z range. If downsampling large cohorts, create replicates at defined sample counts to measure linear or near-linear scaling of runtime and peak memory usage.

Related tools

• ThermoRawFileParser (Convert Thermo .RAW vendor files to mzML format for downstream asari processing) — https://github.com/compomics/ThermoRawFileParser
• ProteoWizard msconvert (Convert most vendor data formats and .mzXML files to centroid mzML) — https://proteowizard.sourceforge.io/tools.shtml
• asari (Process organized centroid mzML files into feature tables; accepts project directory structure created by this skill) — https://github.com/shuzhao-li/asari
• pymzml (Parse mzML files during asari processing; used internally to extract MS1 spectra)

Examples

# Retrieve public test data and organize for asari processing; then profile scaling across sample counts
cd /workspace && wget -r https://github.com/shuzhao-li-lab/data/tree/main/data && ls -la data/*.mzML | head -10 && python3 -m asari.main analyze --input data/sample_001.mzML

Evaluation signals

• All mzML files are centroid-mode (verified via mzML header inspection or asari analyze summary); no profile-mode files present
• Sample registry contains complete metadata (sample ID, ionization mode, file path, acquisition date) for all files
• Directory structure is consistent and navigable; all file paths in registry are valid and resolve without errors
• For cohorts used in scalability benchmarking: runtime and peak memory scale linearly or near-linearly with sample count; 2–3 replicate runs per cohort show mean and standard deviation with <20% relative variability
• mzML files retain full MS1 scan sequence; spot-check a file with pymzml or asari analyze to confirm scan count, m/z range, and retention time span match instrument specifications

Limitations

• Public datasets (MetaboLights, MassIVE) may have variable metadata quality, missing acquisition parameters, or different instrument configurations — manual curation and documentation are required
• Format conversion (e.g. .RAW to mzML) can introduce subtle artifacts or loss of vendor-specific metadata; always validate the converted mzML files against original vendor software if critical quality metrics are needed
• Large cohorts (100+ samples) require substantial disk space (~1–10 GB per sample depending on instrument and run time); plan storage accordingly
• Scalability benchmarking assumes consistent hardware and system load; timing/memory measurements are sensitive to concurrent processes and disk I/O contention

Evidence

• [other] Retrieve public mzML datasets from MetaboLights or MassIVE repositories; select or subsample test data to create cohorts of 10, 50, and 100+ samples.: "Retrieve public mzML datasets from MetaboLights or MassIVE repositories; select or subsample test data to create cohorts of 10, 50, and 100+ samples."
• [readme] Input data are centroid mzML files from LC, GC or DI-MS metabolomics.: "Input data are centroid mzML files from LC, GC or DI-MS metabolomics."
• [readme] We use ThermoRawFileParser to convert Thermo .RAW files to .mzML. Msconvert in ProteoWizard can handle the conversion of most vendor data formats and .mzXML files.: "We use ThermoRawFileParser to convert Thermo .RAW files to .mzML. Msconvert in ProteoWizard can handle the conversion of most vendor data formats and .mzXML files."
• [readme] MS/MS spectra are ignored in default LC-MS workflow but handled by alternative workflows.: "MS/MS spectra are ignored in default LC-MS workflow but handled by alternative workflows."
• [methods] Build sample registry; see workflow.process_project, workflow.register_samples.: "Build sample registry; see workflow.process_project, workflow.register_samples."
• [methods] The default method uses pymzml to parse mzML files.: "The default method uses pymzml to parse mzML files."