FIGURE 1.

Workflow of N‐glycosylation analysis of discovery and validation pancreatic cancer case‐control cohorts for classification analysis. A, Collection of serum samples from pancreatic cancer patients and healthy controls. B, Random distribution of age‐and sex‐matched case‐control pairs, in‐house standards and blanks. C, Automated sample preparation including enzymatic glycan release, derivatization and purification. D, MALDI‐FTICR‐MS analysis of N‐glycome. E, MS‐spectrum preprocessing, annotation and quality control. F, Derived trait calculation for the analysis of glycosylation features. G, Logistic regression analysis of both cohorts, followed by meta‐analysis of the data. H, ROC analysis to test glycosylation traits for their classification power