Related Experiment Video
Updated: May 5, 2026
12:47
Workflow Based on the Combination of Isotopic Tracer Experiments to Investigate Microbial Metabolism of Multiple Nutrient Sources
Published on: January 22, 2018
9.0K
Improving Isotope Ratio Accuracy in Metabolic Labeling Using Orbitrap Mass Spectrometry: A Machine Learning
Zhenwen Yu1, Naveed Ziari1, Marc K Hellerstein1
1Department of Nutritional Sciences & Toxicology at University of California Berkeley, Berkeley, California 94720, United States.
Analytical Chemistry
|May 4, 2026
Summary
A new machine learning method corrects biases in Orbitrap mass spectrometry, significantly improving the accuracy of stable isotope labeling studies for metabolic flux analysis. This advances high-throughput "fluxomics" research.
Area of Science:
- Metabolomics
- Analytical Chemistry
- Systems Biology
Background:
- Stable isotope labeling is crucial for studying metabolic fluxes, with mass spectrometry being the primary measurement tool.
- Orbitrap mass spectrometers offer high accuracy in targeted measurements but exhibit bias in untargeted analyses across wide mass ranges.
- Measurement bias in isotope ratio determination is influenced by factors like ion signal intensity.
Purpose of the Study:
- To develop a machine learning-based correction method for scan-by-scan bias prediction in mass isotopomer ratios.
- To address and overcome the limitations of current Orbitrap mass spectrometers in achieving accurate isotope measurements for untargeted analyses.
Main Methods:
- Development of a random forest machine learning model for predicting bias-free mass isotopomer ratios.
- Scan-by-scan correction applied to address bias in isotopic measurements.
- Identification of bias-causing factors, particularly the ion signal intensity/TIC ratio.
Main Results:
- The machine learning model significantly reduced the mean absolute percentage error for M1 isotopomers (21.3% to 3.5%) and M2 isotopomers (25.8% to 3.1%).
- Ion signal intensity/TIC ratio was identified as the dominant, previously neglected, bias-causing factor.
- Improved accuracy in metabolic flux measurements from heavy water (2H2O) labeling studies.
Conclusions:
- The developed correction method enhances the accuracy of isotope ratio measurements in Orbitrap mass spectrometry for wide mass ranges.
- This approach enables high-throughput stable isotope labeling experiments.
- The method advances the field toward untargeted metabolomics and 'fluxomics'.
