SemiNovo: Learning Beyond Search-Identified Spectra for De Novo Peptide Sequencing
preprint · arXiv · 2026
| Date | 2026-11-16 |
| Type | preprint |
| Venue | arXiv |
| Publisher | arXiv |
| Contribution | algorithm |
| DOI | 10.48550/arXiv.2611.21079 |
Abstract
Sensitive peptide identification is fundamental to protein characterization and novel protein discovery, for which de novo peptide sequencing (DNPS) from tandem mass spectra provides an important database-independent approach. Although deep learning has substantially advanced de novo sequencing, existing models are predominantly trained on high-confidence peptide-spectrum matches obtained through database search. This search-conditioned supervision excludes a large fraction of unidentified spectra, creating a distribution mismatch between the selectively labeled training data and the complete spectrum stream encountered at inference time, thereby limiting peptide identification performance. To address this challenge, we propose SemiNovo, a semi-supervised framework that learns jointly from identified and unlabeled spectra. SemiNovo first constructs DarkSpec, a large-scale quality-controlled corpus of unlabeled spectra collected from diverse public proteomics projects, thereby extending training beyond search-identified spectra and mitigating the resulting shift in DNPS supervision. It then combines stepwise mass-constrained autoregressive decoding with cumulative path confidence to derive reliable supervision from unlabeled spectra. Moreover, SemiNovo introduces head-wise chemical-evidence gating to improve spectral evidence aggregation under noisy and incomplete fragmentation. Experiments show that SemiNovo performs favorably against state-of-the-art methods on multiple benchmarks. Code and dataset are publicly available.
Methods and tools
- SemiNovo: Semi-supervised training on identified and 4.5 million unlabeled spectra (DarkSpec), pseudo-labelled by stepwise mass-constrained autoregressive decoding with cumulative path confidence, plus head-wise chemical-evidence gating of spectral attention.
Data deposited
- DarkSpec (v1) · PanLiu/DarkSpec
Trained on
- DarkSpec (v1) · PanLiu/DarkSpec
- Nine-species benchmark (InstaNovo split) · 10.57967/hf/3821, InstaDeepAI/ms_ninespecies_benchmark
- ProteomeTools (HC-PT (NovoBench)) · no public address
- Seven-species benchmark (NovoBench split) · no public address
Evaluated on
- Nine-species benchmark (InstaNovo split) · 10.57967/hf/3821, InstaDeepAI/ms_ninespecies_benchmark
- ProteomeTools (HC-PT (NovoBench)) · no public address
- Seven-species benchmark (NovoBench split) · no public address