torchcell.metabolism.parameters.resolve_kcat_table

resolve_kcat_table(units: CatalyticUnits, model_dir: str, oed_mirror_dir: str, *, registry: PredictorRegistry | None = None, target_temperature_c: float = 30.0) → ParameterTable

One \(k_{cat}\) per catalytic unit, database-first.

A catalytic unit is an AND-term of genes: a complex has several, an isozyme one. The unit’s turnover is taken as the minimum over its member genes’ resolved values, because a complex cannot turn over faster than its slowest constituent, and this is the same “scarcest subunit limits” logic the availability softmin uses one level up.

Resolution per gene: UniProt accession from SwissProt -> Open Enzyme Database rows for that accession -> the existing temperature-aware selection cascade (resolve_parameter()) -> a registered sequence predictor if one is present -> the organism default. The organism default is the median of whatever the database did supply, not a literature constant, so it is at least on the right scale for the organism at hand; with an empty slice it falls back to FALLBACK_KCAT_PER_S.

Returns:

A ParameterTable of length units.n_units, in units of 1/s.