Seven-species benchmark
benchmark · DDA · 2 versions · 14 papers
| Kind | benchmark |
| Acquisition | DDA |
DeepNovo’s LOW-resolution evaluation set, assembled from seven earlier publications and scored leave-one-out: train on six species, test on the seventh. Distinct data from the high-resolution nine-species benchmark, not a version of it, though both come from Tran et al. 2017 and papers name them side by side.
Versions
original (DeepNovo, 2017)
The low-resolution seven-species set as DeepNovo defined and used it, scored leave-one-out over the seven species. No single accession is published for it: the seven constituent datasets come from seven separate prior publications.
introduced by De novo peptide sequencing by deep learning
No public address: this version is named in the literature but cannot be downloaded.
NovoBench split
NovoBench’s fixed split: yeast held out as the test species and the other six used for training, 3 PTMs, mean peptide length 15.79. This is the split NovoBench retrains every architecture on, so a number reported “on seven-species” by a NovoBench-derived paper means this and not DeepNovo’s own leave-one-out.
train 317,009 · validation 17,740 · test 17,094 · introduced by NovoBench: Benchmarking Deep Learning-based De Novo Peptide Sequencing Methods in Proteomics
No public address: this version is named in the literature but cannot be downloaded.
Deposited by (2)
Each of these introduced a DIFFERENT version, listed beside it; a deposit itself happens once.
- NovoBench: Benchmarking Deep Learning-based De Novo Peptide Sequencing Methods in Proteomics (2024) ML conference NovoBench split
- De novo peptide sequencing by deep learning (2017) peer-reviewed original (DeepNovo, 2017)
Used by (12)
- π-MNovo improves de novo peptide sequencing through microbial-domain adaptation and evidence-guided candidate selection (2026) preprint NovoBench split
- GyroNovo: Error-Guided Fragment Imputation with Mass-Aware Attention for De Novo Peptide Sequencing (2026) preprint NovoBench split
- Discrete Diffusion with Physical Mass Constraints for De Novo Peptide Sequencing (2026) ML conference NovoBench split
- LIPNovo+: Self-Reflective Latent Imputation for Robust De Novo Peptide Sequencing (2026) preprint NovoBench split
- Regressor-guided Diffusion Model for De Novo Peptide Sequencing with Explicit Mass Control (2026) peer-reviewed NovoBench split
- CausalNovo: Advancing De Novo Peptide Sequencing via a Causality-Informed Framework (2026) preprint NovoBench split
- Curriculum Learning for Biological Sequence Prediction: The Case of De Novo Peptide Sequencing (2025) preprint NovoBench split
- Latent Imputation before Prediction: A New Computational Paradigm for De Novo Peptide Sequencing (2025) preprint NovoBench split
- Bridging the Gap between Database Search and De Novo Peptide Sequencing with SearchNovo (2025) ML conference NovoBench split
- ReNovo: Retrieval-Based De Novo Mass Spectrometry Peptide Sequencing (2024) ML conference NovoBench split
- Bridging the Gap between Database Search and De Novo Peptide Sequencing with SearchNovo (2024) preprint NovoBench split
- Enhancing Peptide Mass Spectra Encoder through Pretraining using Contrastive Predictive Coding (2024) thesis NovoBench split
Methods on these papers (13)
- CPC spectrum encoder pretraining adjacent
- CausalNovo algorithm
- DeepNovo algorithm
- DiffuNovo algorithm
- GyroNovo algorithm
- LIPNovo algorithm
- LIPNovo+ algorithm
- NovoBench benchmark
- PhysNovo algorithm
- ReNovo algorithm
- RefineNovo algorithm
- SearchNovo adjacent
- π-MNovo algorithm
Taken from the describing links only, so a paper that merely ran a tool on this data does not make that tool a method of it.