Extraction of species-specific glycan substructures.
Glycans are an important class of biological macromolecules in addition to DNAs and proteins. Actually, glycans play key roles in cellular functions including cell-cell communications, protein interactions and immunity. In this paper, we conducted a comprehensive analysis on comparative glycomics using glycan structures stored in the KEGG/GLYCAN database . First, we developed a new similarity measure for comparing glycan structures and tested its ability to classify glycans of different blood components in the framework of Support Vector Machine (SVM). The result shows that our method successfully classified glycans from different human blood cells. Next, we extracted characteristic functional units (motifs) of glycans suspected to be substructures specific to each blood component. Finally, we conducted an experiment based on agglutination assay in order to verify the result of our prediction, and we confirmed that the fungal lectin specifically recognized the glycan motif predicted by our method.