Data Sources
Last updated: 8 August 2026
Sensorist AI is built on published datasets, regulatory records and the scientific literature. The corpus is curated: raw materials are matched, cross-checked and de-duplicated across sources before anything reaches an answer.
Listed below are the datasets whose licences ask to be credited, with the attribution each one specifies. Answers are grounded in this material; they do not quote any single source verbatim.
FlavorDB2
Flavour molecules and their food associations.
Garg N et al., FlavorDB: a database of flavor molecules, Nucleic Acids Research, gkx957 (2017). IIIT-Delhi.
ChemTastesDB
Curated basic-taste and chemesthetic classifications for molecular tastants.
ChemTastesDB, doi:10.5281/zenodo.5747393. Licensed CC BY 4.0.
M2OR
Human olfactory receptor–odorant activation data, including potency.
Lalis M, Hladiš M, Abi Khalil S, Briand L, Fiorucci S, Topin J. M2OR: a database of olfactory receptor–odorant pairs, Nucleic Acids Research 52, D1370 (2023).
EFSA OpenFoodTox
EU safety assessments of food-relevant substances.
European Food Safety Authority (EFSA), OpenFoodTox. Re-used with acknowledgement per the EFSA legal notice.
EU Union List of Flavourings
Regulation (EC) No 1334/2008 Annex I — FL numbers and the flavouring substances authorised in the EU.
© European Union, via EUR-Lex. Reuse under Commission Decision 2011/833/EU (CC BY 4.0).
JECFA
WHO/FAO Joint Expert Committee safety evaluations of flavouring agents.
Joint FAO/WHO Expert Committee on Food Additives (JECFA) evaluations, World Health Organization.
Beyond these, the corpus draws on public chemical registries, regulatory texts and open scientific literature that carry no attribution condition — among them PubChem and PubMed abstracts from the US National Library of Medicine.
If you maintain one of these resources and want to discuss how your data is represented, contact hello@sensorist.ai.