Uncertainty in protein-ligand binding constants: asymmetric confidence intervals versus standard errors

Area of Science:

• Biochemistry
• Biophysics
• Computational Chemistry

Background:

• Equilibrium binding constants (K_b) quantify molecular interactions, vital for drug discovery and computational biology.
• Current methods often assume symmetric uncertainty distributions for K_b, potentially misrepresenting experimental data.
• Asymmetric distributions in K_b, K_d, and ΔG_b are common but not always appreciated.

Purpose of the Study:

• To illustrate the impact of asymmetric binding constant distributions in isothermal titration calorimetry (ITC) experiments.
• To demonstrate how transforming binding constants (K_b, K_d, ΔG_b) alters their asymmetry.
• To advocate for a standardized and more accurate method for reporting binding uncertainty.

Main Methods:

• Utilized a simulation approach to model asymmetric binding constant distributions.
• Analyzed transformations between K_b, K_d, and ΔG_b to assess changes in asymmetry.
• Examined binding reaction data obtained from ITC experiments.

Main Results:

• Asymmetric distributions significantly affect the interpretation of binding constants derived from ITC.
• Transforming between K_b, K_d, and ΔG_b changes the degree of distribution asymmetry.
• Standard deviation-based uncertainty representation can be misleading for asymmetric distributions.

Conclusions:

• The common practice of representing binding constant uncertainties with symmetric distributions is often inadequate.
• Reporting 95% confidence intervals provides a more accurate and reliable measure of uncertainty for K_b, K_d, and ΔG_b.
• Adopting 95% confidence intervals aligns with best practices and improves data interpretation in molecular interaction studies.