The sense of smell, is initiated by the activation specific receptors expressed at the surface of olfactory neuron, induced by the binding of an odorant. This first step activates signaling pathways leading to the perception of distinct odors. To gain a deeper understanding of the molecular mechanisms involved in olfaction, we have gathered a database of OR-odorant pairs, drawn from published literature and public databases. It includes over 51,483 unique pairs of ORs-molecules. The database contains information about the responses, receptors, molecules, and experiments that have been performed.
Provided identifier in the source are used to find InChI Keys on PubChem (Name, CID, CAS, InChI Key). We use InchI Keys as the standard identifier for the database. In case only the structure is available, the canonical SMILES canonicalSMILES is deduced using openbabel. If the isomerism is not specified, the molecule is considered as “sum of isomers” as long as it has at least one chiral center. The count of chiral centers is determined with RDKit. Mixtures of odorants are identified as "mixture" and written as a space-separated list of InChI Keys (Mixture).
Uniprot IDs are obtained from either the sequence, given identifiers, or when necessary, the name (Gene Name, Uniprot ID). If the Uniprot ID is not available, only the sequence is included (Sequence). In cases where the receptor is mutated, the wild-type identifier or sequence is retrieved, and the mutation is indicated separately using the AApositionAA format (Mutation). The taxon name of the family from which the OR come from is also included (species).
Response is described by a binary code: 0 non-agonist, 1 agonist (Responsive). All decisions regarding responsiveness are made by the original authors, except for Mainland et al. In the case of primary or secondary screening, the concentration used (Value_Screen) and its unit (Unit_Screen) is specified as well as, when available, the raw response value (Value) and its unit (Unit). For dose-response measurements, EC50 value is added in Value, Unit columns. In case the ligand is a non agonist, "n.d" is used to describe the pair as non-responsive and the maximum concentration used in the test is also included. Moreover, the number of independent experiment is included (Nbr_measurements).
The biological test is described by the quantity measured (Type), the OR expression system (*Cell_line, Gprotein, Co_transfection, Tag), the experimental conditions (Delivery, Assay) and finally the tooling (Assay system).
The reference (Reference, DOI) which the pairs originated from is given as well as the location (Reference Position) (Table, Figure, Supp Mat) where the information was found.
@inproceedings{
hladi{\v{s}}2023matching,
title={Matching receptor to odorant with protein language and graph neural networks},
author={Matej Hladi{\v{s}} and Maxence Lalis and Sebastien Fiorucci and J{\'e}r{\'e}mie Topin},
booktitle={The Eleventh International Conference on Learning Representations },
year={2023},
url={https://openreview.net/forum?id=q9VherQJd8_}
}