Search the PubChem PUG REST API by compound name, SMILES, InChI, InChIKey, or molecular formula and return matching PubChem Compound IDs (CIDs).
Usage
pubchem_search(
query,
type = c("name", "smiles", "inchi", "inchikey", "formula"),
cache = TRUE
)Arguments
- query
Character string. The search term (compound name, SMILES, InChI, InChIKey, or molecular formula).
- type
Character string. The type of identifier in
query. One of"name"(default),"smiles","inchi","inchikey", or"formula".- cache
Logical. If
TRUE(default), results are cached in the session environment so repeated searches for the same query and type are instant.
Value
A tibble with a single column cid containing integer PubChem CIDs.
Returns a zero-row tibble if no matches are found.
Details
SMILES and InChI inputs are automatically sent via POST to avoid URL encoding
issues with special characters (/, #, +).
PubChem may return multiple CIDs for a single name query (e.g., different stereoisomers). If more than 10 CIDs are returned, a message is displayed with the count.
For structure-based searches (SMILES, InChI), the query is sent as a POST
form body rather than in the URL path. This is required because
httr2::req_url_path_append() does not percent-encode path segments,
and characters like / and # in SMILES notation would corrupt the URL.
See also
pubchem_properties() to fetch properties for CIDs,
pubchem_synonyms() to fetch synonym lists,
ct_chemical_detail_search() for CompTox Dashboard chemical search.
Examples
if (FALSE) { # \dontrun{
# Search by name
pubchem_search("aspirin")
# Search by SMILES (uses POST automatically)
pubchem_search("CC(=O)OC1=CC=CC=C1C(=O)O", type = "smiles")
# Search by InChIKey
pubchem_search("BSYNRYMUTXBXSQ-UHFFFAOYSA-N", type = "inchikey")
# Search by molecular formula
pubchem_search("C9H8O4", type = "formula")
} # }