Ion mobility calibration curve fitting

ion-mobility-calibration-curve-fitting

Summary

Fit a regression model (linear or polynomial) to drift time vs. collision cross section (CCS) reference standards to establish a calibration curve for converting experimental TWIM-MS arrival times into calibrated CCS values. This is a core step in high-dimensional multi-omic ion mobility mass spectrometry data analysis.

When to use

You have TWIM-MS experimental data with arrival times and m/z values, and access to calibrant reference standards with known CCS values (typically loaded from a calibration template). You need to convert raw arrival times into calibrated CCS values across the m/z range before downstream biomolecular class assignment or quantification.

When NOT to use

• Arrival times have already been manually converted to drift times using a different method or instrument calibration—re-fitting may introduce inconsistency.
• Reference calibrant standards are unavailable, unlabeled, or have uncertain CCS values.
• The experimental m/z range extends far beyond the m/z coverage of reference standards, leading to extrapolation beyond the validated calibration domain.

Inputs

• TWIM-MS arrival times (time at detector) and m/z values from experimental sample
• Instrument-specific time-of-flight offset parameter
• Calibrant reference standards with known CCS values
• Calibration template (data structure defining reference standards and CCS assignments)

Outputs

• Calibration curve model (linear or polynomial regression coefficients)
• Calibrated CCS values for all ions in experimental dataset
• CCS table with m/z, charge state, drift time, and quality metrics
• Calibration quality report (e.g., residuals, R² goodness-of-fit)

How to apply

Load calibrant reference standards with known CCS values from the calibration template. Convert experimental arrival times to drift times by subtracting the instrument-specific time-of-flight offset. Fit a linear or polynomial regression model to the calibrant reference standards, mapping drift time to CCS across the m/z range. Apply the resulting calibration function to the experimental drift times to generate calibrated CCS values for all ions. Export the calibrated CCS table with associated m/z, charge state, and quality metrics (e.g., residuals or goodness-of-fit). Validate that calibration residuals are acceptable and that the curve spans the full m/z range of experimental data.

Related tools

• MOCCal (Python application that implements CCS calibration as a core workflow step, including data import, drift time conversion, calibrant loading, and calibration curve fitting for TWIM-MS data) — https://github.com/HinesLab/MOCCal
• DEIMoS (Dependency for MOCCal when processing raw (unprocessed) calibration files; version 1.3.2 required for RawDT workflow) — http://github.com/pnnl/deimos

Evaluation signals

• Calibration curve residuals (difference between fitted and reference CCS values) are within acceptable bounds (typically < 2% relative error for high-quality standards).
• Goodness-of-fit metric (R² or adjusted R²) is ≥ 0.99, indicating the model explains ≥ 99% of variance in reference standards.
• Calibrated CCS values for unknown samples fall within the m/z and drift time range spanned by the reference standards (no extrapolation).
• Quality metrics (charge state, m/z assignment) are correctly propagated to output CCS table and match input experimental data.
• Calibration is reproducible: re-fitting with the same standards and parameters yields identical or near-identical CCS values (bit-identical or within floating-point tolerance).

Limitations

• TWIM platforms record arrival time (time at detector) rather than drift time (time in TWIM cell), and the conversion requires an accurate instrument-specific time-of-flight offset; incorrect offset propagates systematic error into all calibrated CCS values.
• Linear or polynomial regression assumes a smooth, monotonic relationship between drift time and CCS; non-ideal behavior (e.g., instrument hysteresis or space-charge effects) may violate this assumption and degrade calibration accuracy at extreme m/z or CCS values.
• Reference calibrant standards must cover the m/z and CCS range of experimental data; sparse or unevenly distributed standards lead to poor fit quality or unreliable predictions in sparse regions.
• No changelog provided; version history and known issues are not documented in the repository README.

Evidence

• [other] Convert arrival times to drift times by accounting for the time-of-flight offset specific to the instrument platform.: "Convert arrival times to drift times by accounting for the time-of-flight offset specific to the instrument platform."
• [other] Load calibrant reference standards with known CCS values from the calibration template.: "Load calibrant reference standards with known CCS values from the calibration template."
• [other] Apply linear or polynomial regression to establish a calibration curve mapping drift time to CCS across the m/z range.: "Apply linear or polynomial regression to establish a calibration curve mapping drift time to CCS across the m/z range."
• [other] Apply the calibration function to experimental arrival times to generate calibrated CCS values for all ions.: "Apply the calibration function to experimental arrival times to generate calibrated CCS values for all ions."
• [readme] TWIM platforms record the time at which the ion reaches the detector (arrival time) rather than the time an ion spends within the TWIM cell (drift time).: "TWIM platforms record the time at which the ion reaches the detector (arrival time) rather than the time an ion spends within the TWIM cell (drift time)."
• [readme] Functionality includes collision cross section (CCS) calibration, experimental data biomolecular class assignment, and experimental class-specific CCS calculations.: "Functionality includes collision cross section (CCS) calibration, experimental data biomolecular class assignment, and experimental class-specific CCS calculations."