Method for detecting rare differences between two LC-MS runs

Area of Science:

• Biopharmaceutical analysis
• Analytical chemistry
• Mass spectrometry

Background:

• LC-MS based multi-attribute methods (MAM) are crucial for monitoring biopharmaceutical quality attributes.
• Detecting new or missing peaks is essential for MAM implementation and comparative analysis.
• Comparing similar samples without replicates is challenging due to signal intensity-dependent MS variability.

Purpose of the Study:

• To develop a statistical method for detecting rare differences between two very similar samples.
• To enable reliable comparison of samples without the need for replicate analyses.
• To enhance the accuracy of new-peak detection in biopharmaceutical quality control.

Main Methods:

• A novel statistical approach was developed assuming most components are equivalent between samples.
• The method accounts for signal intensity-dependent relative variability.
• Monoclonal antibody peptide mapping datasets were analyzed to validate the method.

Main Results:

• The statistical method successfully detected rare differences between highly similar samples.
• The approach demonstrated suitability for new-peak detection in MAM.
• A significant reduction in false positive rates was achieved without substantially increasing false negative rates.

Conclusions:

• The developed statistical method offers a robust solution for identifying subtle variations in complex samples.
• This technique enhances the reliability of MAM for biopharmaceutical characterization.
• The method has broader applicability for detecting rare differences in various scientific investigations.