A comparative study of in vitro dose–response estimation under extreme observations

Abstract

Quantifying drug potency, which requires an accurate estimation of dose–response relationship, is essential for drug development in biomedical research and life sciences. However, the standard estimation procedure of the median–effect equation to describe the dose–response curve is vulnerable to extreme observations in common experimental data. To facilitate appropriate statistical inference, many powerful estimation tools have been developed in R, including various dose–response packages based on the nonlinear least squares method with different optimization strategies. Recently, beta regression‐based methods have also been introduced in estimation of the median–effect equation. In theory, they can overcome nonnormality, heteroscedasticity, and asymmetry and accommodate flexible robust frameworks and coefficients penalization. To identify a reliable estimation method(s) to estimate dose–response curves even with extreme observations, we conducted a comparative study to review 14 different tools in R and examine their robustness and efficiency via Monte Carlo simulation under a list of comprehensive scenarios. The simulation results demonstrate that penalized beta regression using the mgcv package outperforms other methods in terms of stable, accurate estimation, and reliable uncertainty quantification.

REAP-2: An interactive quantitative tool for robust and efficient dose-response curve estimation

“…The innovation not only ameliorates accuracy for the interested estimands, such as Hill coefficients and drug potency in both point and interval estimation, but also improves efficiency with narrower confidence intervals at the same nominal level. Most importantly, it largely enhances reliability of robust beta regression framework in dose-response estimation, despite various patterns of extreme data in real-life applications [6]. The results in previous comparative study demonstrate that the updated REAP-2 provides reliable estimations and possesses higher power in statistical testing.…”

“…Furthermore, to better assist the dose-response estimation, we conducted a comparative study [6] using a Monte Carlo simulation to review 14 different dose-response estimation tools including dose-response R packages and beta regression-based algorithms. We examined them in 30 different scenarios with various settings of extreme responses and concluded that in general, the penalized beta regression using the mgcv R package [7] had the best performance in terms of feasibility, accuracy, and coverage probabilities of the target estimands for doseresponse.…”