Skip to main content
. 2021 Nov 25;18(12):1515–1523. doi: 10.1038/s41592-021-01306-0

Fig. 3. O-glycopeptide identification and SSGL of pGlyco3.

Fig. 3

ac, Software comparisons of O-glycopeptide searches with IHMO HEK-293 cell line data. a, The overlaps of other tools with pGlyco3 on O-GPSMs. For glycans, only the total glycan compositions were compared; SSGL was not considered. ETD scans in the results of Byonic were mapped to their corresponding HCD scans for the comparisons. If an HCD spectrum and its sister ETD spectrum were identified as the same glycopeptide, we kept only one GPSM; otherwise, we kept both. b, The identified IHMO O-GPSMs were validated by Hex-containing results and further validated by Hex-diagnostic ions. c, The runtime comparison. d, Validation of the SSGL-FDR of pGlycoSite using the entrapment-based SSGL-FDR and OpeRATOR-based SSGL-FDR (Methods). e, Localized site-specific O-glycans of ITIH4, KNG1 and F12 proteins in human serum samples. Site-groups were discarded and SSGL assignments with maximal probability ≥0.75 are displayed. f, An annotated spectrum of the localized O-glycan and its SSGL probability for KNG1-S403. The HCD spectrum is annotated in Supplementary Fig. 11. The ScoreTable with BestPath is shown in Supplementary Data.