simbiology-fit-model

Use fitproblem instead of calling sbiofit directly. Set the Model, Data as groupedData, ResponseMap in "Species = DataColumn" format, and Estimated as an estimatedInfo object with bounds, then call fit(prob) to run the estimation.

sbiofit handles single-subject or pooled fitting and enforces parameter bounds, while sbiofitmixed performs population NLME fitting with inter-individual variability and random effects but ignores bounds. Select via the FitFunction property on fitproblem.

Use sbionca with an sbioncaoptions object specifying concentrationColumnName, timeColumnName, and a dose column such as EVDoseColumnName. Metrics like AUC_0_last, C_max, T_half, and CL are returned, with NaN required on non-dosing rows.

Common causes are missing bounds on estimatedInfo, not using log transforms for rate constants spanning orders of magnitude, or wrapping data in a plain table instead of groupedData. Always set Bounds and IndependentVariableName before fitting.

No, sbioparameterci only supports results from sbiofit nonlinear regression, not sbiofitmixed NLME results. For sbiofit results, use Gaussian or ProfileLikelihood types to compute parameter confidence intervals.

Use sbiosampleparameters with the CovariateModel expression, fixed effects, and random effect covariance matrix from NLME results to sample parameter sets. Then simulate them with createSimFunction to produce virtual patient cohorts.