Affinity Determination: A Practical Guide to Measuring Molecular Binding Strength (KD, KA, kon, koff)

What “Affinity Determination” Means

Affinity determination is the process of quantifying how strongly two molecules bind to each other—commonly protein–protein, antibody–antigen, receptor–ligand, or protein–small molecule interactions. In most bioscience and drug discovery contexts, affinity is summarized by the equilibrium dissociation constant (KD):

• Lower KD = higher affinity (tighter binding).

• KD is an equilibrium quantity, meaning it reflects the balance between binding and unbinding at steady state.

A related way to express the same concept is the association constant (KA), where KA = 1 / KD. 

The Core Parameters: KD, KA, kon, koff

Affinity can be described in two complementary ways:

1) Equilibrium view (how much binds at steady state)

• KD (M): concentration at which half of binding sites are occupied in a simple 1:1 interaction model.

• KA (M⁻¹): binding strength as an association constant (inverse of KD). 

2) Kinetic view (how fast binding happens)

Many instruments determine affinity by measuring rates:

• kon (M⁻¹·s⁻¹): association/on-rate (how quickly complex forms)

• koff (s⁻¹): dissociation/off-rate (how quickly complex falls apart)

For a 1:1 interaction:

KD = koff / kon (at equilibrium). Surface-based biosensors often estimate affinity by extracting these rates from real-time binding curves. 

Why Affinity Determination Matters (Beyond “Strong vs Weak”)

Affinity is not just a bragging-rights number. It influences:

• Dose requirements (tighter binding can enable lower effective concentrations)

• Specificity and selectivity (high affinity can help, but does not guarantee specificity)

• Residence time (often tied to koff; slow off-rate can matter in pharmacology)

• Comparability across candidates (rank-ordering binders in discovery pipelines)

Main Experimental Strategies for Affinity Determination

Different methods trade off realism, throughput, equipment needs, and what they measure (equilibrium vs kinetics).

A) Surface Plasmon Resonance (SPR): Real-Time Kinetics + KD

SPR measures binding in real time by detecting refractive index changes near a sensor surface. It is widely used for extracting kon, koff, and KD from sensorgrams (binding curves). Key ideas include reference subtraction, controlling immobilization, and fitting to binding models (often starting with a 1:1 Langmuir model). 

Strengths

• Provides kinetics (kon/koff) and affinity (KD) in a single workflow

• Label-free detection

Common pitfalls

• Mass transport limitations (apparent kinetics distorted by diffusion limits)

• Surface immobilization artifacts (orientation, density, steric hindrance)

• Non-specific binding and baseline drift that complicate fitting 

B) Isothermal Titration Calorimetry (ITC): Direct Thermodynamics + KA/KD

ITC measures heat released/absorbed during binding in solution. From titration curves, it can yield:

• KA (or KD)

• ΔH (enthalpy)

• ΔS (entropy)

• n (stoichiometry)

This is powerful because it connects “how strong” with “why strong” (enthalpy-driven vs entropy-driven binding). 

Strengths

• Fully solution-based, label-free

• Rich thermodynamic interpretation

Common pitfalls

• Requires sufficient heat signal and careful concentration design

• Extremely tight binding can require special designs (e.g., displacement approaches) 

C) Competition ELISA: Accessible Affinity Estimation in Solution

Competition ELISA can estimate binding strength by letting soluble analyte compete with immobilized antigen (or vice versa). Under proper conditions, it can approximate true solution affinity without specialized instrumentation, and is often used when SPR/ITC is unavailable. 

Strengths

• Lower barrier to entry (standard immunoassay equipment)

• Useful for comparative ranking when carefully designed 

Common pitfalls

• Assay conditions must preserve equilibrium assumptions

• Immobilization and avidity (multivalency) can skew apparent affinity

• IC50-to-KD relationships require correct regime and modeling 

Data Interpretation: “True Affinity” vs Apparent Affinity

A frequent source of confusion in affinity determination is apparent affinity—a value that looks like KD but actually reflects additional effects:

• Avidity (multivalent interactions can appear tighter than single-site binding)

• Heterogeneous binding sites (mixed populations produce non-1:1 behavior)

• Surface artifacts (for SPR/ELISA formats)

A robust affinity workflow typically includes:

• Testing multiple concentrations

• Checking residuals and model fit quality

• Running controls for non-specific binding

• Evaluating whether a simple 1:1 model is justified 

Choosing the Right Affinity Determination Method (Rule-of-Thumb)

• If you need kon/koff + KD and can work with surfaces: SPR 

• If you need thermodynamics (ΔH/ΔS) + stoichiometry in solution: ITC 

• If you need accessible, instrument-light comparisons: competition ELISA 

In practice, teams often combine methods: for example, SPR for kinetic ranking plus ITC on finalists for mechanistic confidence.