This function estimates a collection of growth parameters from the mean GP fits. If necessary, it also runs the GP fits.
Usage
estimate_growth_parameters(
object,
model_covariate = "curve_id",
predict_n_steps = 100,
downsample_every_n_timepoints = 1,
sample_posterior_gpfit = FALSE,
od_auc_at_t = NULL,
sample_n_curves = 100,
save_gp_data = FALSE,
n_cores = 1,
detect_polyauxic = FALSE,
polyauxic_ratio_threshold = 0.2,
polyauxic_parameter_criteria = "carrying_capacity"
)
# S4 method for class 'DGrowthR'
estimate_growth_parameters(
object,
model_covariate = "curve_id",
predict_n_steps = 100,
downsample_every_n_timepoints = 1,
sample_posterior_gpfit = FALSE,
od_auc_at_t = NULL,
sample_n_curves = 100,
save_gp_data = FALSE,
n_cores = 1,
detect_polyauxic = FALSE,
polyauxic_ratio_threshold = 0.2,
polyauxic_parameter_criteria = "carrying_capacity"
)Arguments
- object
A DGrowthR object containing preprocessed data.
- model_covariate
A string indicating a covariate in metadata to pool growth curves for GP modelling.
- predict_n_steps
A numeric value indicating the number of timepoints to sample from the mean posterior of the GP fit.
- downsample_every_n_timepoints
A numeric value indicating that the OD from every n timepoint should be used for GP fit. Might seep up fitting.
- sample_posterior_gpfit
A logical value. Indicates if the posterior GP should be sampled for new growth curves and parameters are estimated from these sampled curves. Allows to estimate mean and standard deviation of growth parameters. Really only makes sense if model_covariate pools more than one growth curve per group.
- od_auc_at_t
A numeric value indicating a specific timepoint for which the predicted OD and AUC should be returned.
- sample_n_curves
A numeric values. If sample_posterior_gpfit is TRUE, then sample_n_curves are sampled from posterior.
- save_gp_data
A logical value, indicating if the mean GP values and GP fit parameters should be saved to object.
- n_cores
A numeric values indicating the number of cores the user wants to use to model curves in parallel.
- detect_polyauxic
A logical value indicating if the function should attempt to detect polyauxic growth curves. This is done by looking for multiple peaks in the second derivative of the mean GP fit. If TRUE, then additional columns to the growth parameter table will be added with the relevant information.
- polyauxic_ratio_threshold
A numeric value. Potential growth phases are assessed by comparing their respective carrying capacity. If the carrying capacity of a potential second phase is at least this fraction of the carrying capacity of the first phase, then the second phase is considered a true growth phase. Only relevant if detect_polyauxic is TRUE.
- polyauxic_parameter_criteria
A character value. Either "growth_rate" or "carrying_capacity". Indicates which parameter should be used to assess the relevance of potential growth phases. Only relevant if detect_polyauxic is TRUE.