MALDI Deamidation Score (MDS): A fast and flexible method for assessing deamidation in ZooMS data and its application to the Denisova Cave bone assemblage

Abstract

Estimating deamidation from ZooMS spectra has frequently been achieved using the q2e method due to its high-throughput capacity and ease of use. Despite this, q2e only works with spectral data in txt format, operates slowly, and employs a genetic algorithm for fitting, which is stochastic and less interpretable. Furthermore, q2e only estimates deamidation at the peptide level and does not generate a sample-wide summary. The introduction of the Parchment Glutamine Index (PQI) presented an alternative method. Initially designed for a large ZooMS dataset of parchment, it utilises weighted least squares and a linear mixed-effects model (LME) to generate a peptide deamidation estimation and a sample level index, respectively. To address the limitations of q2e and expand the applicability of PQI to a wider range of archaeological tissues and MALDI-TOF-derived data (such as ZooMS data on bone collagen), we developed the MALDI Deamidation Score (MDS), an iteration of the PQI method optimised for handling large-scale datasets. MDS is more streamlined for analysing multi-species data with customisable peptide lists, offering dramatically reduced processing time. Using the published Denisova Cave ZooMS assemblage, we demonstrate that different peptides exhibit varying deamidation patterns over time, making the use of a single peptide to represent overall deamidation potentially biased. Such information is invaluable for investigating key questions such as protein preservation and site formation processes.