dIon-de-novo-labeled-v1 (DNL-v1)
training · DDA · 1 version · 1 paper
dIon’s supervised de novo training corpus, combining MassIVE-KB, ProteomeTools and further labelled sources, split by canonical peptide identity with exact-spectrum duplicates removed. Internally called KitchenSink V5.
| Kind | training |
| Acquisition | DDA |
| Home | https://huggingface.co/datasets/alfred-n/dIon-datasets/tree/main/de_novo/corpora/DNL-v1 |
dIon’s supervised de novo training corpus, combining MassIVE-KB, ProteomeTools and further labelled sources, split by canonical peptide identity with exact-spectrum duplicates removed. Internally called KitchenSink V5.
Versions
v1
train 5,275,467 · validation 170,140 · test 345,414 · released 2026-10-09 · introduced by dIon: Fragmentation-Based Invariance for Self-Supervised Learning of Tandem Mass Spectra
Where it lives:
- Hugging Face · alfred-n/dIon-datasets#de_novo/corpora/DNL-v1 de_novo/corpora/DNL-v1
Deposited by (1)
Methods on these papers (1)
- dIon 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.