
NMR Online Limited // Manchester, UK
Biomolecular NMR analysis, in a browser.
Processing to model-free. Assignment to Δχ tensors. Twenty-nine analysis capabilities and 111 executable workflows, wrapping the tools your field already trusts — with every result traceable back to the FID it came from.

- 29
- analysis capabilities
- 111
- executable workflows
- 10
- engines built in-house
- 100%
- results with provenance
Nugget Run
A guinea pig, a hedge maze full of nuggets, and coyotes that would rather you did not finish. Eat a big nugget to become a capybara, and the coyotes become lunch. Arrow keys or WASD; space to pause.
Loading the maze…
One toolchain, reachable two ways
A seasoned spectroscopist configures a pipeline directly through the forms. Someone newer describes what they want in words. Both drive the same workflow engine and write to the same result store — so there is one implementation, one set of results, and one audit trail.
The assistant is not a separate product bolted on the side. Every analysis service already publishes an OpenAPI specification, and the assistant's tools are generated from those specifications — so its capabilities track the platform's without anyone wiring them up twice.
Correctness does not rest on the model. A deterministic guardrail, derived from the platform's own workflow definitions, gates every operation on the artifacts it requires actually existing. Fitting before assignment transfer, or model-free before a correlation time, is structurally impossible rather than merely discouraged.
The scientist drives, the assistant does the legwork — picking, integrating and assigning peaks, driving the viewer, and launching the right pipeline — and the scientist verifies every result. Consequential output arrives marked proposed, for you to accept or discard.
Every number traces back to the FID
Derived NMR parameters — S², kex, Kd, Δχ tensors — are routinely published without a machine-readable trail from raw data to result. Deposition standards solve data exchange. They do not solve analysis provenance.
Every result is stored as an envelope recording the inputs it consumed and the outputs it produced. Any value on the right of that graph can be walked back to the spectrum and structure it came from — months later, by someone else, without asking you what you did.
Eight engines we built ourselves
Wrapping the community's tools is table stakes. These are ours — written because the existing options could not answer the question, or could not say how confident they were.
CHORD
Backbone dihedrals from shifts, with calibrated confidence
Cartouche
Probabilistic assignment that tells you which experiment to run next
Cadence
Model-free, spectral densities, diffusion tensors and detectors
PICA
Peak picking and cluster deconvolution, 1D to 3D
pcs-match
PCS assignment transfer with simultaneous Δχ-tensor fitting
DWAB
CEST pre-fit that removes the initialisation failure mode
Titration
Peak matching and per-residue Kd fitting across a series
Planner
Tells you an experiment will not work before you book the magnet
What you can run
Twenty-nine capabilities across nine domains, each backed by one or more validated engines.
Processing & reconstruction
Peak picking & fitting
Assignment
Relaxation & dynamics
Chemical exchange
RDC & paramagnetic
Binding & titration
Prediction & validation
Pipelines & automation
We are community led
Our online NMR software is engineered with scientists firmly in mind. We actively seek community engagement, to ensure we engineer the NMR software that spectroscopists actually want… and will be able to use.
Focus on the science
We engineer web NMR software that is intuitive and visually distinctive, developed using modern software industry principles. Not only agile, efficient and up-to-date — rigorously tested and fully supported, too.
Learn more
NMR Online Limited has a professional and friendly reputation. Follow us on social media for the latest news, or get in touch and tell us what you are trying to measure.
Start with the free tier
ELEMENTS covers the core analyses at no cost. ELECTRO adds data reuse across every calculation, the Data Explorer, projects, per-residue results, custom pipelines and the experiment planner.