Modanovo: A Unified Model for Post-translational Modification-Aware De Novo Sequencing Using Experimental Spectra...

Area of Science:

• Proteomics
• Computational Biology
• Mass Spectrometry

Background:

• Post-translational modifications (PTMs) are crucial for cellular regulation and disease.
• Database searching for modified peptides is limited by combinatorial complexity.
• Existing de novo sequencing methods struggle with PTMs.

Purpose of the Study:

• To develop a de novo peptide sequencing method capable of identifying peptides with diverse PTMs.
• To enhance the comprehensiveness of proteomic analysis by including modified peptides.
• To provide a complementary approach to database searching for peptide identification.

Main Methods:

• Curated a large dataset of spectra for 19 amino acid-PTM combinations (phosphorylation, acetylation, ubiquitination).
• Developed Modanovo, an extension of the Casanovo transformer architecture for de novo sequencing.
• Evaluated Modanovo's performance against existing tools and database search methods.

Main Results:

• Modanovo achieved high performance on modified peptides (AUC 0.92) and unmodified peptides (AUC 0.93).
• Outperformed other de novo sequencing tools (π-PrimeNovo-PTM, InstaNovo-P) and showed complementarity with MSFragger.
• Identified novel viral phosphosites in a monkeypox virus dataset not found by database search.

Conclusions:

• Modanovo is a robust and broadly applicable tool for de novo sequencing of both modified and unmodified peptides.
• Demonstrates significant potential for advancing proteomic research, especially in PTM discovery.
• Complements traditional database search methods, enabling more comprehensive peptide identification.