Converting IC₅₀ to \(K_d\)#
An IC₅₀ is an assay-dependent midpoint, whereas \(K_d\) is an equilibrium
dissociation constant. bindcurve implements three conversions for the
specific case of one-site, mutually exclusive, competitive equilibrium between
a labeled tracer and an unlabeled competitor.
The notation is:
Symbol |
Code |
Meaning |
|---|---|---|
\([R_T]\) |
|
total receptor concentration |
\([L_T^*]\) |
|
total tracer concentration |
\(K_d^*\) |
|
tracer dissociation constant |
\(K_d\) |
|
competitor dissociation constant |
\(\mathrm{IC}_{50}\) |
|
total competitor giving 50% tracer displacement |
Here, 50% displacement means \([RL^*]_{50}=[RL^*]_0/2\), where the subscript 0 denotes the competitor-free state. These conversions do not apply to noncompetitive or irreversible mechanisms, nonequilibrium measurements, functional-response IC₅₀ values, heterogeneous sites, or the four-state incomplete-competition model.
Cheng-Prusoff approximation#
Function: cheng_prusoff
The familiar conversion used by the API is
The classical relationship is written in terms of free tracer. Substitution of
total tracer is accurate when tracer depletion by receptor is negligible, so
\([L^*]\approx[L_T^*]\), and bound competitor contributes negligibly to its total
concentration. In bindcurve, this approximation is the low-receptor limit of
the finite-concentration conversions below.
Munson-Rodbard finite-concentration correction#
Function: cheng_prusoff_corrected
Let
be the bound-to-free tracer ratio before competitor is added. Under the same one-site competitive-equilibrium model, the exact finite-concentration correction is
The second term is subtracted. The original article printed a plus sign; the authors’ erratum corrected it to a minus sign. When \(y_0\to0\), this expression reduces to the Cheng-Prusoff approximation.
Some combinations of \(\mathrm{IC}_{50}\), \([L_T^*]\), \(K_d^*\), and \(y_0\) are
incompatible with the assumed equilibrium and produce a non-positive result.
bindcurve rejects such inputs rather than reporting them as affinities.
Nikolovska-Coleska finite-concentration correction#
Function: coleska
This form derives the competitor-free state from \([R_T]\), \([L_T^*]\), and \(K_d^*\) instead of requiring \(y_0\).
Before competitor is added, free receptor obeys
where
The physical root is
The remaining competitor-free quantities are
At 50% displacement,
and tracer equilibrium gives
Receptor conservation then requires
Both terms on the right are subtracted. The free competitor concentration is
and the competitor dissociation constant follows directly from its equilibrium:
An algebraically equivalent final expression used by the implementation is
Under consistent inputs, the Munson-Rodbard and Nikolovska-Coleska corrections recover the same one-site equilibrium \(K_d\). The former requires \(y_0\); the latter derives the initial state from total receptor and tracer concentrations.
References#
Cheng, Y. & Prusoff, W. H. (1973), Relationship between the inhibition constant and the concentration of inhibitor which causes 50 per cent inhibition.
Munson, P. J. & Rodbard, D. (1988), An exact correction to the Cheng-Prusoff correction, with the published erratum.
Nikolovska-Coleska, Z. et al. (2004), Development and optimization of a binding assay for the XIAP BIR3 domain using fluorescence polarization.