Comparative analysis of glycoproteomic software using a tailored glycan database.

Area of Science:

• Glycoscience
• Proteomics
• Bioinformatics

Background:

• Glycoproteomics is rapidly advancing, driven by technological innovations.
• Numerous software tools exist for glycoproteomic data analysis, making comparison challenging.
• Standardized comparison of these tools is needed to guide researchers.

Purpose of the Study:

• To conduct a head-to-head comparison of five modern glycoproteomic analysis software packages.
• To evaluate software performance using matched glycan databases derived from sample glycomic profiling.
• To identify the most reliable software for glycopeptide spectrum identification and analysis.

Main Methods:

• Compared Byonic, Protein Prospector, MSFraggerGlyco, pGlyco3, and GlycoDecipher.
• Minimized parameter variables and constructed matched glycan databases for each software.
• Analyzed glycopeptide spectra from three replicates of wild-type SH-SY5Y cells, identifying up to 17,000 spectra per software.

Main Results:

• Significant overlap was observed across software for identified glycoproteins, glycosite locations, and glycans.
• No single software demonstrated superior performance across all criteria.
• Evidence suggests Byonic may report spurious results at glycoprotein and glycosite levels.

Conclusions:

• Assessing glycoproteomic software requires multiple comparative criteria beyond just the number of spectra identified.
• Glycoproteomic data analysis should involve consensus from multiple software tools, excluding the current version of Byonic, for increased confidence.
• Future research could explore combining peptide-first and glycan-first analytical approaches.