Research On Glycan Structure Identification Algorithm Based On Mass Spectrometry

Glycans are very important organic macromolecules in nature,and they play a very crucial role in biological processes.Different from the simple linear structures of nucleic acids and proteins,branching patterns of glycan are more complicated,and there is no gene template for glycans synthesis,making their research extremely slowly.With the development of mass spectrometry(MS),tandem mass spectrometry(MS~2)has provided favorable tools for the research of glycans,but it usually cannot accurately achieve the identification of branching patterns of glycan.Recently,multi-stage mass spectrometry(MS~n)is being applied,but there is not a scheme for glycan patterns identification which could efficiently guide MS~n experiments.In response to above problems,the main contents of this thesis are as follows:(1)In response to Electrospray Ionization(ESI)MS,complemented Matrix-assisted Laser Desorption Ionization(MALDI)MS,we developed the MS~nspectra annotation algorithm,called gly Branch,in an ESI version,which improved the applicability of gly Branch.We proposed an effective pre-processing for spectra of ESI-MS~n,and successfully used gly Branch to accurately analyze the spectra of standard glycan samples using ESI-MS~n.We confirmed that gly Branch could fully use MS~ninsights and report the results of identification.(2)We achieved a best first search algorithm,called best FSA,based on MALDI-MS~n.There are two parts in best FSA.Firstly,best FSA constructs a scoring function based the tree of spectra using MS~n,which could stably and conveniently score the candidate branching patterns of glycans.Secondly,by using the best first search strategy,best FSA selects the node with minimum distance on the tree of spectra using MS~n,efficiently and accurately achieving the identification of branching patterns of glycans.In response to the problems,consisting of that MS~2 is not enough to identify the branching patterns of glycans and a lack of efficient scheme for MS~n experiments,this thesis achieved a best first search algorithm based MALDI-MS~n,which could accurately complete the identification of branching patterns of glycan,reducing the cost and complexity of MSn experiments,helping researchers have a more convenient and efficient insight in fragmentation of glycans.