Data completeness scores
Stars on nmrXiv compounds measure how complete the publicly available NMR evidence is for a compound's constitution. They are not a measure of spectral quality, spectrometer field strength, or whether the proposed structure is correct.
Tiers
| Stars | Label | Requirements |
|---|---|---|
| 0 | Not yet rated | No public spectra scored for this compound |
| 1 | Spectra available | At least one public spectrum |
| 2 | 1D characterised | ¹H and ¹³C |
| 3 | Heteronuclear correlation | ¹H, ¹³C, and HSQC or HMBC |
| 4 | Full elucidation set | ¹H, ¹³C, COSY/TOCSY, HSQC, and HMBC |
| 5 | Fully assigned | Full elucidation set plus NMR assignments |
Bonus badges (shown when present, not required for any tier):
- DEPT — carbon multiplicity
- NOESY/ROESY — through-space contacts for relative configuration
See Why each experiment matters for the chemistry behind these choices.
What counts
- Only public studies and datasets are scored.
- Experiment types come from NMRium spectrum metadata (
nucleus,experiment,dimension). - Assignments count when either:
- the Assignments tab has an ACS-style string or atom–peak table, or
- NMRium ranges/zones contain non-empty
diaIDslinking atoms to signals.
- Every score stores which rubric version produced it. When the scoring rules change, scores are recomputed.
Where you see stars
- Compound cards in search and library results
- The compound page checklist (present / missing experiments)
- The compound library data-quality panel (contributor / workspace stars)
FAQ
My compound has fewer stars than expected
Check that the missing experiments are published and that NMRium recognises their experiment type (for example hsqc, hmbc, cosy). Private drafts do not count.
Why didn't my DEPT count as ¹³C?
DEPT does not observe quaternary carbons, so it cannot replace a ¹³C spectrum. DEPT is shown as a bonus badge instead. A multiplicity-edited HSQC often covers the same multiplicity information once ¹³C and HSQC are present.
How do I raise a score?
Add the missing experiment (or assignments), then publish the study. Scores refresh when a study is published and again on a nightly safety-net job.