Ashley Scott

Strategies for data validation in ancient protein studies using a milk model

Ashley Scott§, Jessica Hendy§,£, Christina Warinner§,£.

§ Max Planck Institute for the Science of Human History, Jena, Germany; £ University of Oklahoma, Norman, United States

Technological advances in LC-MS/MS technology are increasingly allowing archaeological researchers to gain new insights into past cultures through the analysis of ancient protein residues. This presentation will focus on how optimization of aspects of study design, quality control, and data authentication can improve data reliability and archaeological interpretations from ancient protein remains. The most common focus of quality control and data validation in the field of ancient proteomics is in the final output, whereby data are generally processed using probabilistic scoring algorithms where significance values are given to help us assess the likelihood that results are valid. However, there are additional checks and authentication steps that can be incorporated into the workflow, even prior to sampling, to improve the reliability and confidence of ancient protein identification. Using archaeological milk proteins as a model system, this presentation will focus on how database design, patterns of degradation (e.g., deamidation) and proteotypic peptide recovery, and other factors can be used to assist in protein identification, validation, and authentication. Additionally, critical attention to the possible limitations of our methods or proteins of interest will be addressed. Validation procedures can be time-consuming but they are a vital component of experimental design and should be a routine part of any ancient proteomics workflow.