π-HelixNovo
algorithm · Transformer (AR)
π-HelixNovo: algorithm · Transformer (AR). Complementary spectra
Complementary spectra
| Kind | algorithm |
| Deep learning | yes |
| Acquisition | DDA |
| Also known as | PandaNovo |
| Family | Transformer (AR) |
Code
Live stars, open issues and last-push figures are on the Code activity chart.
Checkpoints
| Version | Trained on | Host | Licence | Size | Checked |
|---|---|---|---|---|---|
| project archive v2 | — | Zenodo archival | CC-BY-4.0 | 1.6 GB | live 2026-10-02 |
| project archive | — | Zenodo archival | CC-BY-4.0 | 22.6 GB | live 2026-10-02 |
Benchmarks
- denovo_benchmarks: median peptide-level average precision 0.825 over 86 datasets, median rank 4 of 14 (version bm-1.1.0).
- ProteoBench, on the nine-species benchmark, ProteoBench selection: peptide-level AUC 0.626; precision 0.595 at 100% coverage; amino-acid AUC 0.740 (version 1.0, beam search, submitted 2026-08-07).
Both are mass-based matches on the tool’s most recent run. What these numbers mean.
Papers describing it (2)
- Introducing PandaNovo for practical large-scale de novo peptide sequencing (2023, bioRxiv, preprint)
- Introducing π-HelixNovo for practical large-scale de novo peptide sequencing (2024, Briefings in Bioinformatics, peer-reviewed)