Quantitative characterization of hormone receptors

Area of Science:

• Biochemistry
• Pharmacology
• Statistical Modeling

Background:

• Traditional methods like Scatchard plots and linear regression for characterizing hormone receptor binding are statistically suboptimal.
• These methods violate assumptions of linear regression, leading to inaccurate parameter estimations.

Purpose of the Study:

• To introduce and advocate for the use of weighted nonlinear least-squares regression for accurate estimation of receptor binding parameters.
• To highlight the limitations of Scatchard plots and simple linear regression in receptor binding analysis.

Main Methods:

• Utilizing weighted nonlinear least-squares regression with total ligand concentration as the independent variable.
• Employing curve fitting for nonlinear Scatchard plots to determine affinity (K) and capacity (R) for multiple receptor classes.
• Implementing statistical criteria for model selection and assessing goodness-of-fit.

Main Results:

• Weighted nonlinear least-squares regression provides more statistically sound estimates of K and R compared to linear methods.
• The developed computer programs successfully analyze steroid receptor binding in breast carcinoma specimens.
• Alternative techniques like limiting slopes and affinity distributions address complex binding scenarios.

Conclusions:

• Weighted nonlinear least-squares regression is the superior method for analyzing hormone receptor binding data.
• Accurate characterization of receptor binding parameters is crucial for understanding biological processes and drug development.
• Advanced statistical approaches enhance the reliability of receptor binding studies.