Post-search validation of site-resolved N-glycopeptide assigments using complementary LC–MS/MS evidence

Mass spectrometry-based glycoproteomics enables site-resolved characterization of protein glycosylation. However, glycan isomerism, isobaric species, analytical artifacts, and incomplete fragmentation can compromise automated identification. We developed a post-search workflow that validates and curates MS2-based candidate assignments using complementary precursor-level and chromatographic evidence.
Search-engine-derived candidates were reviewed and quantified across runs in Skyline using precursor monoisotope assignment, mass accuracy, isotopic-envelope agreement, fragment-ion evidence, extracted-ion chromatograms, retention-time consistency, and glycosite context. This review refuted inconsistent assignments, removed redundant peptide forms, added MS1-supported composition-level features, and retained LC-resolved isomeric features without assigning specific structures.
Using data from 52 human serum samples, we evaluated 3,071 candidate N-glycopeptide assignments and refuted 1,722 (56%) because they conflicted with precursor level and/or chromatographic evidence. Final curation also incorporated 22 additional MS2-derived entries, removed 255 redundant peptide forms, and added 320 features, comprising 152 MS1-supported composition-level assignments and 168 additional LC-resolved isomeric features. This resulted in 1,436 curated N‑glycopeptide features spanning 148 N-glycosylation sites on 58 glycoproteins.
These results demonstrate that automated MS2 scoring alone is insufficient for reliable large-scale N-glycoproteomics. Integrating precursor, fragment-ion, retention-time, and cross-run evidence improves both assignment accuracy and analytical coverage.