A unified computational ecosystem for authenticated metabolomics — from raw data to published results.
MS/MS matching against an authenticated reference collection. Accurate mass, retention time and spectral similarity are checked independently to deliver MSI Level 1a confidence.
Explore →Query known bioactivity against your identified compounds, cross-referencing KEGG, PubChem, ChEMBL and Reactome to surface plausible targets and pathways.
Explore →Reveal structural relationships between named and unnamed features, so an identified compound lends its context to the dark matter around it.
Explore →Scale identification across whole cohorts. A RESTful endpoint with a Python client for automated, high-throughput processing against the same library.
Explore →Predict retention times for compounds outside the library using models trained on the measured set, giving you an RT prior where no standard exists yet.
Move between KEGG, PubChem CID, ChEMBL ID, InChIKey and SMILES in a single step, so identities stay stable across the tools you already use.
Principal component analysis and correlation structure across your samples, so group separation is visible before you test for it.
Volcano plots with fold-change and significance, backed by Welch’s t-test and ANOVA across your defined sample groups.
Hierarchical clustering, dendrograms, k-means and heatmaps to expose the patterns that separate your conditions.
PLS-DA, random forest and support vector machines for supervised separation, with permutation testing to guard against overfitting.
Spectral similarity networks that place identified and unidentified features in structural context together.
Map identified compounds onto biological pathways to move from a metabolite list to a mechanistic interpretation.
Retention-time drift correction and batch-effect correction across runs, so a multi-day study does not fracture into per-batch artefacts masquerading as biology.
Automated spectral quality assessment flags the acquisitions that are not fit to draw conclusions from, before they reach your results table.
Search a spectrum of interest against public repository data to find where else that molecule has been observed — useful context for a feature your own library cannot yet name.