Vera Ignjatovic,University of Melbourne, Australia
Glycoproteomics holds immense potential to advance understanding of the molecular and cellular processes underpinning human (patho)physiology, but remains an analytically challenging discipline (Thaysen-Andersen M et al., Mol Cell Proteomics. 15(6):1773. 2016). Accurate automated intact glycopeptide identification from high resolution tandem mass spectrometry data, a prerequisite for the further advancement of glycoproteomics, is still in its infancy despite many exciting glycoinformatics developments (Hu et al., Mass Spectrom Rev. 36(4):475. 2017).
Formed in September 2016, the first community study of the B/D-HPP Human Glycoproteomics Initiative (HGI) sets out to provide a detailed comparison of the bioinformatics solutions currently in use for the site-specific determination of protein N- and O-linked glycosylation. For this purpose, two high resolution CID-, HCD- and EThcD-based LC-MS/MS glycopeptide datasets were acquired from a tryptic digest of serum glycoproteins and these two common data files were distributed to all study participants. The participating teams were asked to return lists of confidently identified N- and O-glycopeptides from the two LC-MS/MS data files using preformed reporting templates and, further, to provide information of their used method of identification. Significant efforts by the HGI study committee were invested to ensure uniformity and compliance of all teams with the study guidelines.
An update of the study progress was presented by Dr Thaysen-Andersen at the B/D-HPP Investigator Meeting during the 17th World Congress HUPO, Orlando, FL. The study has been very well received by the community with 22 glycoproteomics laboratories encompassing 9 software developers and 13 user teams located across all major continents completing the identification task by the end of the reporting deadline (October 2018). Recent analysis of the glycopeptide data reports by the HGI study committee encouragingly shows the existing of diverse informatics tools and approaches with a great potential for (semi-)automated glycopeptide identification from complex LC-MS/MS datasets, but, at the same time, highlighted a significant variance of glycopeptide/protein identifications within and between the user and developer teams. Discordance was also observed between participants using the same software. The underlying reasons for this spread and further data comparisons to delineate the more exact overlap and differences between the glycopeptides reported by the 22 teams are currently being investigated.
We aim to publish the 1st HGI study results in 2019 and will present the final study outcomes at the 2nd Australasian Glycoscience Symposium (2nd AGS) to be held on the 14th September in Adelaide, Australia (more information to follow) immediately prior to the 18th World Congress HUPO, Adelaide, Australia where the prospects and scope of a potential 2nd HGI study also will be discussed.