GECS

GECS predicts glycan structures by linking glycosyltransferase (GT) gene expression to the chemical structures of biosynthetic glycans to infer glycan chains.

Key Features:

• Missing-glycan estimation: Estimates missing glycans using a global glycan structure map to infer structures absent from existing databases.
• Real-valued scoring scheme: Scores candidate glycan structures using real-valued gene expression intensities rather than binary conversion.
• Novel candidate discovery: Identifies new glycan structures not present in current databases by leveraging the global glycan structure map and expression links.
• Validated on leukemia datasets: Applied to patient gene expression profiles from acute lymphocytic leukemia (ALL) and acute myeloid leukemia (AML) and shown to achieve statistically significant performance improvements.

Scientific Applications:

• Glycan biosynthesis research: Infers glycan biosynthetic relationships from GT expression to study biosynthesis mechanisms.
• Cancer glycomics: Predicts glycan structures associated with ALL and AML for study of cancer-associated glycans.
• Biomarker discovery and personalized medicine: Aids identification of glycan biomarkers for diagnosis and treatment strategies based on individual gene expression profiles.

Methodology:

Links glycosyltransferase (GT) gene expression to glycan chemical structures, estimates missing glycans via a global glycan structure map, and ranks candidates using a real-valued gene expression intensity scoring scheme applied to patient gene expression profiles.