prior.control {clinDR}

Set the parameters of the prior distribution for the Emax model implemented in fitEmaxB.

Description

Set the parameters of the prior distribution for the Emax model implemented in fitEmaxB. prior.control is deprecated. See emaxPrior.control.

Usage

prior.control(epmu = NULL, epsd = NULL, emaxmu = NULL, 
	emaxsd = NULL, p50 = NULL, 
	sigmalow = NULL, sigmaup = NULL, 
	led50mu = 0.79, led50sca = 0.6, edDF = 3, 
	lama = 3.03, lamb = 18.15, lamsca = 6, 
	basemu=NULL,basevar=NULL,
	binary = FALSE)

Arguments

epmu

Mean for E0 in a normal prior distribution. Logistic scale for binary data.

epsd

SD for E0 in a normal prior distribution. Logistic scale for binary data.

emaxmu

Mean for Emax in a normal prior distribution. Logistic scale for binary data.

emaxsd

SD for Emax in a normal prior distribution. Logistic scale for binary data.

p50

Projected ED50. See reference for its use in creating the prior distribution for the ED50.

sigmalow

Lower bound for a uniform prior distribution for the residual SD (continuous data).

sigmaup

Upper bound for a uniform prior distribution for the residual SD (continuous data).

led50mu

Mean of log-t prior distribution for the ED50 before final scaling. See reference for its interpretation in the prior distribution for the ED50.

led50sca

Scale (analogous to SD) of the log-t prior distribution for the ED50.

edDF

The degrees of freedom of the log-t prior distribution for the ED50.

lama

Parameter in the re-scaled beta distribution for Hill slope parameter in the sigmoidal Emax model. See reference for it use and empirical basis.

lamb

Parameter in the re-scaled beta distribution for Hill slope parameter in the sigmoidal Emax model.

lamsca

The beta prior distribution for the Hill parameter is re-scaled to have support on (0,lamsca).

basemu

A vector of prior means for the covariate regression parameters.

basevar

The prior variance-covariance matrix for the covariate regression parameters. The covariate regression parameters are apriori independent of the other dose response model parameters.

binary

Set to TRUE for binary data applications. Used to check for consistency in usage.

Details

The prior distributions are based two meta-analyses of dose response described in the references. Each parameter is independent in the prior distribution. The E0 and Emax parameters have normal prior distributions. For binary data, these parameters are computed on the logistic scale. The predicted ED50 must be specified as 'P50'. The prior distribution of the log(ED50) has a t-distribution centered at log(P50), with scale, degrees of freedom, and offset to the P50, defaulting to values given in the references (these can be changed, but they are difficult to interpret outside the context of the meta-analyses). If modType=4, the prior distribution for the Hill parameter is a beta distribution scaled to (0,lamsca). The default degrees of freedom were obtained from the meta-analyses. For continuous data, the prior distribution for the residual SD is uniform on a user-specifed scale.

Value

List of prior parameter values for use in fitEmaxB.

Author(s)

Neal Thomas

References

Thomas, N., Sweeney, K., and Somayaji, V. (2014). Meta-analysis of clinical dose response in a large drug development portfolio, Statistics in Biopharmaceutical Research, Vol. 6, No.4, 302-317. <doi:10.1080/19466315.2014.924876>

Thomas, N., and Roy, D. (2016). Analysis of clinical dose-response in small-molecule drug development: 2009-2014. Statistics in Biopharmaceutical Research, Vol. 6, No.4, 302-317 <doi:10.1080/19466315.2016.1256229>

See Also

fitEmaxB