From left: From left: Tay Shi Jie, Dr Ian Walsh and Dr Terry Nguyen-Khuong
Ian Walsh 1, Terry Nguyen-Khuong 1, Katherine Wongtrakul-Kish 1, Shi Jie Tay 1, Daniel Chew 1, Tasha Jose 2, Christopher H. Taron 2 and Pauline M Rudd 1
1 Bioprocessing Technology institute, Agency for Science, Technology and Research (A*STAR), Singapore
2 New England Biolabs
Published in Bioinformatics 2019 35(4): 688-690 (Online Version)
Glycans are carbohydrates that are covalently linked to many glycoproteins. Glycans are attributed to numerous physiological and pathological processes, such as cancer progression  due to the added complexity they impart upon protein structure (Figure 1A). This complexity in glycan structure also has direct consequences for biotherapeutic design and understanding their mechanistic role will illuminate novel targets for drug design .
An ongoing research endeavour of the Analytics group at the Bioprocessing Technology Institute is to develop glyco-analytical pipelines to structurally characterize glycans comprehensively. One of the critical bottlenecks is data analysis, where interpreting large datasets from LC-MS is time consuming and at times complex.
GlycanAnalyzer, is software developed to automate the interpretation of the LC-MS data sets so as to identify glycans in a complex sample. At its core are machine learning and pattern matching algorithms that can monitor glycan changes upon treatment with glycan specific enzymes known as exoglycosidases (Figure 1B). Intensive computations involve pattern-matching many thousands of data points and therefore calculations take place on high performance computer servers. GlycanAnalyzer is able to structurally characterise glycans in 20 minutes, in contrast to manual characterisation that can potentially take days/weeks to complete. It is very accurate, identifying glycans attached to an in-house biotherapeutic with 100% accuracy, compared to 81% using a commercially available software.
The GlycanAnalyzer project was conducted in collaboration with New England Biolabs (U.S.A.) and its user interface can be accessed from standard web browsers such as Chrome/Firefox at the address https://glycananalyzer.neb.com (Figure 1B).
 Doherty, M., E. Theodoratou, I. Walsh, B. Adamczyk, H. Stöckmann, F. Agakov, M. Timofeeva, I. Trbojević-Akmačić, F. Vučković, and F. Duffy, Plasma N-glycans in colorectal cancer risk. Scientific reports, 2018. 8.
 Zhang, P., S. Woen, T. Wang, B. Liau, S. Zhao, C. Chen, Y. Yang, Z. Song, M.R. Wormald, and C. Yu, Challenges of glycosylation analysis and control: an integrated approach to producing optimal and consistent therapeutic drugs. Drug Discovery Today, 2016. 21(5): p. 740-765.
Figure 1. Complexity of glycan structure and the GlycanAnalyzer software to aid in structure interpretation.
(A) Glycan structures are oligosaccharides with complex topologies, isomerism, and monosaccharide composition. Characterising this complexity using conventional LC-MS approaches can be difficult. Enzymes that cleave specific monosaccharides (and linkages) on the glycan can be used and the additional LC-MS information can help sequence the glycan structures. (B) the web interface to the GlycanAnalyzer software is simple and extensive tutorial and help pages are available. The software uses machine learning algorithms to pattern match peak shifts in the LC and MS after application of exoglycosidase enzymes.
Please refer to here for more information on the Analytics Group.