Comparative Evaluation of Glycoproteomics Software for Rare Glycopeptide Identification

Abstract

Advancements in glycoproteomics software have improved glycopeptide identification; however, algorithm differences cause discrepancies in identified glycopeptide, even when identical datasets. We compared five state-of-the-art glycoproteomics software programs (Byonic, MSFragger-Glyco, pGlyco3, Glyco-Decipher, and GRable), investigating their unique capabilities, and examined their ability to identify rare sialic acid-containing glycopeptides (NeuGc and KDN) derived from BJAB-K20 cells, which lack UDP-N-acetylglucosamine 2-epimerase, the rate-limiting enzyme for sialic acid synthesis. Approximately half of the identified glycopeptides were unique to individual tools. Byonic identified the highest number of glycopeptides, whereas Glyco-Decipher and GRable identified complex highly branched glycan structures. NeuGc- and KDN-containing glycopeptides were identified by specific programs, highlighting their capability to handle rare glycan structures. To assess the reliability of these identifications, we reanalyzed the MS/MS spectra for the presence of diagnostic ions corresponding to each identified glycopeptide. Some software programs identified glycopeptides without detecting the corresponding diagnostic ions, raising concerns regarding result reliability. However, leveraging the distinct capabilities of each software enabled us to achieve a comprehensive and reliable analysis of glycopeptides, including those with rare glycan structures. Combining multiple glycoproteomics software programs with complementary strengths and incorporating post- verification steps, such as diagnostic ion analysis, enhances the accuracy and depth of glycopeptide identification.