An automatic peak finding method for LC-MS data using Gaussian second derivative filtering.

Area of Science:

• Analytical Chemistry
• Chromatography
• Mass Spectrometry

Background:

• Pharmaceutical analysis requires robust methods for impurity detection.
• Tracking chromatographic peaks is essential during method development.
• Existing methods may lack automation and adaptability.

Purpose of the Study:

• To develop a highly automated procedure for peak localization and characterization in LC-MS data.
• To address the need for efficient peak detection and tracking in pharmaceutical impurity analysis.
• To create an adaptable algorithm for diverse chromatographic datasets.

Main Methods:

• Development of a digital filter-based algorithm.
• Automatic adaptation of filter settings to the dataset.
• Evaluation using synthetic data with varying signal-to-noise (S/N) ratios, peak widths, and baseline properties.
• Validation with experimental LC-MS datasets.

Main Results:

• Achieved 94% detection of simulated analytical peaks with no false positives, even at S/N=10 and high baseline variability.
• Accurately estimated peak heights and widths within 10% error for 94% and 91% of simulated peaks, respectively.
• Demonstrated excellent agreement between processed reconstructions and raw analytical signals for experimental data, improving visualization.

Conclusions:

• The developed automated procedure offers high accuracy and reliability for peak analysis in LC-MS.
• The algorithm's adaptability makes it suitable for various analytical challenges, including pharmaceutical impurity profiling.
• The method enhances data visualization and simplifies chromatographic peak characterization.