hxsparse: A Feature Extraction Algorithm for Hydrogen Exchange-Mass Spectrometry Based on a Linear Deuteration Model...

Area of Science:

• Analytical Chemistry
• Biochemistry
• Computational Biology

Background:

• Accurate mass extraction of deuterated peptides from LC-MS data in hydrogen exchange-mass spectrometry (HX-MS) is crucial but challenging.
• Traditional LC-MS feature extraction algorithms are not optimized for the complexities of HX-MS data.
• Existing commercial software lacks dedicated, literature-described algorithms for HX-MS specific feature extraction.

Purpose of the Study:

• To develop and describe a novel algorithm, hxsparse, for automated feature extraction in HX-MS data.
• To model peptide deuteration using discrete states and LASSO regularization.
• To evaluate the performance of hxsparse against commercial software and expert curation.

Main Methods:

• Developed hxsparse, an algorithm utilizing an exact model for peptide deuteration states.
• Applied least absolute shrinkage and selection operator (LASSO) regularization to introduce sparsity.
• Compared hxsparse performance with commercial software and manual expert curation on LC-MS data.

Main Results:

• hxsparse successfully extracts average masses of deuterated peptides from HX-MS data.
• Automated feature extraction using hxsparse demonstrated superior performance compared to commercial software.
• hxsparse results closely approximated those achieved by expert manual curation.

Conclusions:

• hxsparse provides an effective and accurate automated solution for feature extraction in HX-MS data analysis.
• The algorithm's performance suggests it is a valuable tool for researchers working with deuterium labeling experiments.
• This work addresses a significant gap in the literature for specialized HX-MS data analysis tools.