DiffNovo: A Transformer-Diffusion Model for De Novo Peptide Sequencing
peer-reviewed · Bioinformatics and Computational Biology (BICOB 2025) · 2025
| Date | 2025-07-24 |
| Type | peer-reviewed |
| Venue | Bioinformatics and Computational Biology (BICOB 2025) |
| Publisher | Bioinformatics and Computational Biology: 17th International Conference |
| Contribution | algorithm |
| DOI | 10.1007/978-3-031-94039-2_8 |
| Citations (OpenAlex) | 0 |
Abstract
Accurate identification of peptides from Data-Independent Acquisition (DIA) mass spectrometry data is crucial for discovering novel peptides and proteins. In this study, we introduce DiffNovo, which integrates transformer-based architectures with a diffusion model to enhance de novo peptide sequencing for DIA data. The experimental results demonstrate that DiffNovo outperforms existing state-of-the-art models in recall and precision for de novo peptide identification. DiffNovo achieves an average improvement of 20.7% in amino acid precision and an increase of 57.2% in peptide precision across various datasets. These substantial gains highlight the efficacy of DiffNovo in delivering accurate peptide sequencing from DIA mass spectrometry data.
Methods and tools
- DiffNovo: Transformer-diffusion model
Data used
- De novo sequencing of DIA data (as deposited) · MSV000082368
Cites (8)
- A transformer model for de novo sequencing of data independent acquisition mass spectrometry data (2025) crossref
- A transformer model for de novo sequencing of data independent acquisition mass spectrometry data (2024) crossref
- Transformer-Based De Novo Peptide Sequencing for Data-Independent Acquisition Mass Spectrometry (2023) crossref
- Accurate de novo peptide sequencing using fully convolutional neural networks (2023) crossref
- BiATNovo: A Self-Attention based Bidirectional Peptide Sequencing Method (2023) crossref
- Deep Active Learning for De Novo Peptide Sequencing from Data-independent-acquisition Mass Spectrometry (2022) crossref
- Deep learning enables de novo peptide sequencing from data-independent-acquisition mass spectrometry (2018) crossref
- De novo peptide sequencing by deep learning (2017) crossref