Fast parametric time warping of peak lists

Area of Science:

• Analytical Chemistry
• Biochemistry
• Data Science

Background:

• Peak alignment is crucial for chromatographic data analysis in fields like metabolomics and proteomics.
• Existing methods, including parametric time warping (PTW), face challenges with complexity and computational time.
• Overfitting is a concern in peak alignment, necessitating robust algorithms.

Purpose of the Study:

• To introduce a novel formulation of Parametric Time Warping (PTW) that operates on peak-picked features.
• To enhance the speed and integration capabilities of PTW within existing analytical pipelines.
• To demonstrate the efficacy of the new PTW approach using real-world datasets.

Main Methods:

• Developed a new PTW formulation that processes peak lists instead of full chromatographic profiles.
• Implemented the updated PTW algorithm in the R package 'ptw' (version 1.9.1+).
• Validated the method on publicly available LC-DAD (grape samples) and LC-MS (apple extracts) datasets.

Main Results:

• The new PTW formulation significantly accelerates the alignment process by orders of magnitude.
• Processing peak features allows for smoother integration into existing data analysis workflows.
• The method demonstrated successful peak alignment on diverse LC-DAD and LC-MS data.

Conclusions:

• The feature-based PTW offers a substantial improvement in speed and usability for chromatographic data alignment.
• This approach effectively addresses the challenges of peak alignment in metabolomics and proteomics.
• The enhanced PTW algorithm provides a powerful tool for researchers in relevant analytical fields.