Ancient Protein Identification and Mass Spectrometry Data Analysis

The aim of this MPhil study was to develop novel models and software tools for the analysis of mass-spectrometric data from degraded and ancient proteins. On the basis of background study in ancient collagen and relevant identification approaches, problems of fossil bone collagen identification were discussed. As a solution, the database named UniColl was designed as a repository of theoretical sequences generated from the known type I collagen sequences. The principle of UniColl was to contain a large number of collagen peptide sequences which can be theoretically produced under certain chemical and mathematical algorithm, to include all the known sequence variation in each peptide.

UniColl has been established and evaluated in this work. As the result, large amounts of theoretical sequence have been generated to cover as much possible collagen sequence variations as we can get based on the known information. The practical utility and quality results of this database was tested with groups of collagen sequences identified for several unknown ancient samples.