Bruker mass spectrum data import
Use when you have received a Bruker Solarix FT-ICR raw data directory (.d format, containing CompassXtract output or native ser/fid transients) and need to import it into a Python-based analysis workflow for FT-MS signal processing, calibration, or molecular formula search.From its SKILL.md
npx -y skills add HolobiomicsLab/asb-skill-collections --skill bruker-mass-spectrum-data-importAssembled from the repository path, not quoted from the project. Check it against their README if it does not work.
One thing to look at
- 15 stars15 stars. Stars are a popularity signal and not a quality one, but at this level it is likely that nobody has read this closely except its author, and you would be relying on your own review.
What its file declares
Copied from the file, not written here
The file declares its own license as CC-BY-4.0. That is the author’s claim about this one file, and it is not the same thing as the license GitHub reports for the repository, which is listed with the other numbers below.
SKILL.md
6.0 KB, 995 tokens by cl100k_base, as published. Nobody here has run it
bruker-mass-spectrum-data-import
Summary
Load and initialize Bruker Solarix FT-ICR mass spectrometry raw data (.d format) into CoreMS data structures, preparing transient or magnitude-mode spectra for subsequent noise thresholding, peak picking, and molecular formula assignment.
When to use
You have received a Bruker Solarix FT-ICR raw data directory (.d format, containing CompassXtract output or native ser/fid transients) and need to import it into a Python-based analysis workflow for FT-MS signal processing, calibration, or molecular formula search. This skill is the mandatory first step when working with Bruker instrument output.
When NOT to use
- Input is already a centroid mass list (CSV, Excel, or mzML) — use a generic mass list importer instead.
- Input is from a different vendor (Thermo, Spectroswiss, MagLab) — use vendor-specific readers.
- The .d directory is corrupted or missing mandatory calibration metadata — validate file integrity first.
Inputs
- Bruker Solarix .d directory (containing ser, fid, or CompassXtract output files)
- MSParameters object with field-strength calibration constants
Outputs
- CoreMS MassSpectrum object with m/z, intensity, and metadata
- Transient or magnitude-mode frequency-domain spectrum ready for noise and peak processing
How to apply
Use CoreMS ReadBrukerSolarix to load the .d directory into a CoreMS mass spectrum object. Initialize MSParameters with field-strength-specific settings (e.g., 12 T or 15 T magnetic field calibration constants). Set noise thresholding method (relative_abundance, log, or signal_noise) and peak prominence thresholds (e.g., peak_min_prominence_percent = 1, noise_threshold_min_relative_abundance = 1) according to instrument resolution and expected dynamic range. The importer automatically handles FT magnitude-mode conversion and transient metadata extraction. Verify that m/z range, resolving power, and baseline noise floor are consistent with acquisition parameters before proceeding to peak picking.
Related tools
- CoreMS (Provides ReadBrukerSolarix importer and MassSpectrum data structure for Bruker .d file parsing and FT-ICR metadata extraction) — https://github.com/EMSL-Computing/CoreMS
- Bruker Solarix (Source instrument and data acquisition platform generating .d directories with ser/fid transient or CompassXtract processed spectra)
Examples
from corems.transient.input.brukerSolarix import ReadBrukerSolarix; ms = ReadBrukerSolarix('path/to/sample.d'); print(ms.number_of_peaks, ms.resolving_power_calculator())
Evaluation signals
- MassSpectrum object is non-null and contains non-zero intensity array with matching m/z length.
- Metadata fields (scan_number, rt, polarity, resolving_power, field_strength) are populated and consistent with acquisition header.
- m/z range and mass resolution match Bruker instrument specifications (e.g., 50–1000 m/z at >100k resolving power for 12 T).
- Intensity baseline and dynamic range are appropriate for FT magnitude mode (no truncation or overflow).
- No exceptions or file-read errors are raised; ReadBrukerSolarix completes without timeout.
Limitations
- Only FT magnitude mode is supported; phase-corrected time-domain data requires external preprocessing.
- CompassXtract output may lose transient metadata; ser/fid native import is more reliable for advanced signal processing.
- Bruker proprietary binary formats are reverse-engineered; future CompassXtract versions may cause read failures.
- Import performance degrades for very large transients (>1 GB); streaming mode not yet available.
Evidence
- [other] Load the recalibrated FT-ICR mass spectrum data file (Bruker .d format) using CoreMS ReadBrukerSolarix.: "Load the recalibrated FT-ICR mass spectrum data file (Bruker .d format) using CoreMS ReadBrukerSolarix."
- [other] Import and process transient data is step 1 in FT-MS workflow.: "1. Import and process transient data"
- [readme] CoreMS supports direct access for almost all vendors' data formats.: "CoreMS supports direct access for almost all vendors' data formats"
- [readme] Bruker Solarix transients, ser and fid (FT magnitude mode only) are supported input formats.: "Bruker Solarix transients, ser and fid (FT magnitude mode only)"
- [other] Initialize mass spectrum parameters for a 12 T field-strength instrument, setting noise thresholding method.: "Initialize mass spectrum parameters for a 12 T field-strength instrument, setting noise thresholding method (e.g., relative_abundance or log) and peak prominence thresholds."
What ships with it
Read from the repository
Just SKILL.md. No reference files, no scripts.