I'd like to ask if I'm running a nonlinear mixed effect model for repeated measures of biomarker or echocardiographic measurements, how I can choose the nonlinear equation.