Glycan-binding proteins (GBPs) are valuable tools in glycan analysis because of their high specificities for particular motifs. Glycan arrays offer the possibility of unprecedented insights into the binding preferences of GBPs, but the interpretation of glycan-array data often requires expert knowledge and significant time owing to the complex rules governing GBP binding. Here we introduce a glycan-array analysis platform called MotifFinder that provides the automated determination of details about binding specificities. The method builds models of GBP binding through multivariate regression-trees that incorporate an algorithm for mining flexible and discontinuous motifs. The method is designed to identify not only the primary determinants of GBPs, but also the fine-specificity motifs that modulate gradations in binding. In tests across datasets from multiple, highly-disparate lectins, MotifFinder consistently provided accurate results and identified binding determinants not readily apparent through manual analysis. The system can provide additional insights into GBP binding through the simultaneous analysis of multiple datasets differing in glycan content and lectin concentration. Furthermore, we demonstrated that the quantitative, statistical output from the program has value in downstream applications. The output from the automated MotifFinder analyses could be used to interpret patterns of lectin binding to a glycoprotein in order to uncover information about the protein’s glycosylation. We anticipate that the automated glycan-array analysis system will enable researchers from all disciplines to create and work with accurate models of GBP specificities.

Automated identification of gradations in determinant fine-specificities from glycan array data

2019 Spring National Meeting

The official archive and repository for all ACS Meetings & Expositions abstracts and select posters or presentations.