Designing methods and software to suit the needs of O-glycopeptide characterization Nicholas M. Riley(1), Lei Lu(3), Stacy Malaker(1), Michael R. Shortreed(3), Lloyd M. Smith(3), Carolyn R. Bertozzi(1,2)* (1) Department of Chemistry and Stanford ChEM-H, Stanford University, Stanford, California, USA (2) Howard Hughes Medical Institute, Stanford, California, USA (3) Department of Chemistry, University of Wisconsin, Madison, WI 53706 Mucin-type O-glycosylation is a prevalent post-translational modification on extracellular and secreted proteins that drives both biochemical and biophysical interactions at the cell surface. O-glycoproteins containing heavily glycosylated mucin domains affect immune signaling and host-pathogen interactions, and they have been under scrutiny for years due to their association with cancer. O-glycosylated sequences are heterogeneous both in the glycosites that are occupied and the glycans that modify them, requiring site-specific characterization to connect glycosylation state with biological effect. Tandem mass spectrometry (MS) is the premier method to map O-glycosites, but O-glycopeptides are difficult to properly characterize using standard (N-)glycoproteomic approaches. Here we present data comparing collision-and electron-based tandem MS dissociation methods to underscore the necessity of electron-based fragmentation for O-glycosite mapping. Furthermore, we describe a new database searching strategy that we designed with these lessons in mind. Our so-called O-Pair Search uses paired collision- and electron-based dissociation spectra collected for the same precursor ion to 1) identify O-glycopeptides through an ion-indexed open modification search and 2) localize O-glycosites using graph theory and probability-based localization. O-Pair Search reduces search times more than 2,000-fold compared to current O-glycopeptide processing software (a major challenge for large-scale O-glycoproteomics), while generating ~50% more O-glycopeptide identifications. O-Pair Search also reports localization levels that indicate if all (Level 1), at least one (Level 2), or none (Level 3) of the O-glycosites are confidently localized – a new and much needed paradigm for glycoproteomics. These improvements represent an important shift toward more confident O-glycopeptide characterization while enabling proteome-scale O-glycosite mapping. Additionally, O-Pair Search is easily accessible through the free and open-source MetaMorpheus platform.