Subscribe Bookmark RSS Feed



Aug 19, 2014

Crossed Curves Fitting: A JMP® Application to Characterise a Drug Activity Using Integrated Nonlinear Models

Paul Fogel, Independent Statistical Consultant

Jonathan Pascau, HTS Team Leader, Galderma R&D

David Piwnica, Head of In Vitro Pharmacology, Galderma R&D

Angélique Féret, HTS HQ Tech, Galderma R&D

Assessing the mechanism of action of agonist and antagonist chemical compounds – whether surmountable or insurmountable, allosteric, partial, etc., is key to the research and development of new drugs in the pharmaceutical industry. Typically, multiple dose-response curves of a given reference compound are obtained at different doses of the test compound. Until most recently, each curve used to be fitted independently in order to analyse the type of dependency between usual estimators (EC50s) and the dose of the test compound. Thanks to the availability of computational resources and user-friendly software interfaces, integrated models, which analyse all curves at once, can now provide robust estimates of the parameters of interest (KB, EC50, confidence intervals, etc.) with the additional benefit of ensuring that the estimation process is repeatable and not operator-dependent. The JMP scripting language (JSL) offers the possibility to design user-friendly dialogs, a powerful nonlinear platform and interactive graphics that appear quite useful whenever outliers need to be excluded, for instance. Moreover, our script allows for easy import/export connections with standard databases.