Model

Pydantic models for PubChem assay and compound-property payloads.

DEFAULT_PROPERTIES = ['SMILES', 'XLogP', 'ExactMass', 'MolecularWeight', 'Charge', 'Complexity', 'TPSA', 'InChIKey']

AffinityStrategy

Bases: StrEnum

An enum to specify how to aggregate multiple activity values for the same CID within an assay and same activity type.

ALL = 'all'

AVERAGE = 'average'

MIN = 'min'

AssayInfo

Bases: BaseModel

A parsed PubChem concise assay record.

active_cids

CIDs marked Active in the assay.

aid

The PubChem Assay ID.

inactive_cids

CIDs marked Inactive in the assay.

model_config = ConfigDict(populate_by_name=True)

pubmed_id

Return the PMID shared by all records, or None if they disagree or none is set. Per-row PMIDs live on record.pubmed_id.

records = Field(default_factory=list)

One record per row of the concise table.

unique_cids

All distinct, non-blank CIDs in the assay (any outcome).

binding_affinity(cids=None, strategy=AffinityStrategy.ALL, outcome=None)

Map CID to a list of (activity_value, activity_name, outcome) entries.

PARAMETER	DESCRIPTION
cids	Optional iterable of CIDs to keep. If None, all are kept. TYPE: Iterable[int | str] | None DEFAULT: None
strategy	"all" keeps every value; "min"/"average" collapse to one entry per activity_name using non-None values. TYPE: AffinityStrategy | str DEFAULT: ALL
outcome	Optional filter for activity outcome (e.g. "active"). If None, all outcomes excluding unspecified are kept. TYPE: str | None DEFAULT: None

RETURNS	DESCRIPTION
dict[int, list[tuple[float | None, str, str]]]	dict mapping CID to a list of affinity value_or_None, activity_name and outcome tuples

cids(outcome)

CIDs for an arbitrary outcome (e.g. 'inconclusive', 'probe').

from_concise(aid, data)

Parse a concise assay payload into an AssayInfo.

PARAMETER	DESCRIPTION
aid	The PubChem Assay ID the payload belongs to. TYPE: int
data	The raw concise JSON response (already confirmed to be a dict without a Fault by the client). TYPE: dict[str, Any]

RETURNS	DESCRIPTION
AssayInfo	An AssayInfo. Records/pubmed_id are empty/None if the table is absent.

to_dataframe(outcome=None)

Return the flattened rows as a pandas DataFrame, optionally filtered.

to_rows(outcome=None)

Flatten to one dict per record. Optionally filter by outcome.

CompoundProperties

Bases: BaseModel

Compound properties for a single CID from the property endpoint.

charge = Field(default=None, alias='Charge')

Net formal charge of the molecule.

complexity = Field(default=None, alias='Complexity')

Bertz/Hendrickson/Ihlenfeldt molecular complexity rating. Smaller values indicate simpler molecules.

exact_mass = Field(default=None, alias='ExactMass')

Mass of the most likely isotopic composition (mass-spec oriented).

inchikey = Field(default='', alias='InChIKey')

Hashed InChI; canonical 27-character compound identifier.

model_config = ConfigDict(populate_by_name=True, extra='allow')

molecular_weight = Field(default=None, alias='MolecularWeight')

A sum of all atomic weights of the atoms in the compound, measured in g/mol.

smiles = Field(default='', alias='SMILES')

SMILES string including stereochemistry and isotopes.

tpsa = Field(default=None, alias='TPSA')

Topological polar surface area.

xlogp = Field(default=None, alias='XLogP')

Computed octanol-water partition coefficient (hydrophobicity).