 |
 |
Hello, I have been searching for two days now for the following question and could not find what I was looking for so I hope I just didn't miss something on the blog. my question is this: I have a model where a response is a functional form of other parameters and a covariate (specifically a 4 parameter Emax model for dose response). We are interested in predicting at what dose we will achieve a specific placebo corrected response: RESPONSE - E0 = value. Note that the 4-parameter Emax model is RESPONSE = E0 - (Emax - E0) / (1 + [ED50/DOSE]^gamma) when we solve for the required dose we get DOSE = ED50 / gamma'th root of [(Emax - E0)/(RESPONSE - E0) - 1] as you can immediately see, there are possible situation in which the node DOSE may not be defined when looking for a specific response; that is the radical is negative or zero. Using WinBUGS I obtain a trap error for these situations and since WinBUGS does not handle if then statements I am not sure how to assign this particular iteration of the DOSE variable to be either missing, NaN, NA, or some other way to determine that for the given MCMC observations of E0, Emax, ED50, and gamma that DOSE dose not exist. We would be interested in tracking the probability that this DOSE dose not exist. This can be easily handled if we export the coda and perform such observations in SAS, however, how might we handle this situation in WinBUGS itself interactively? I thought about specifically testing to see if the radical would be negative and assign these to be zero using the code PCR = Placebo Corrected Response of interest for which we want to know what dose gives us this PCR denom <- (Emax-E0)/(PCR) - 1 test <- step(denom)*denom DOSE <- ED50 / pow(test, 1/gamma) but then the question remains how do I get a posterior summary from a variable where we exclude certain responses (zero in this case) from this summary? thoughts? Thank you very much for the time. Have a wonderful day, William H. Atkinson, Ph.D Biostatistician Biostatistics PPD, Austin Texas Phone +1-512-747-5196 Fax +1-512-685-9454 [log in to unmask] www.ppdi.com <http://www.ppdi.com/> This email transmission and any documents, files or previous email messages attached to it may contain information that is confidential or legally privileged. If you are not the intended recipient or a person responsible for delivering this transmission to the intended recipient, you are hereby notified that you must not read this transmission and that any disclosure, copying, printing, distribution or use of this transmission is strictly prohibited. If you have received this transmission in error, please immediately notify the sender by telephone or return email and delete the original transmission and its attachments without reading or saving in any manner. ------------------------------------------------------------------- This list is for discussion of modelling issues and the BUGS software. For help with crashes and error messages, first mail [log in to unmask] To mail the BUGS list, mail to [log in to unmask] Before mailing, please check the archive at www.jiscmail.ac.uk/lists/bugs.html Please do not mail attachments to the list. To leave the BUGS list, send LEAVE BUGS to [log in to unmask] If this fails, mail [log in to unmask], NOT the whole list