Back to the Top
Hi everyone...
I have some question if you can help me;)
First...it's about bootstrap in NONMEM... I want to know what is the
percentual of accepted minimization successful? 80%? And if I have
some minimization with error 134 that I can easily solve, changing the
sicdigits, can I count this one as successful minimization?
Second, it's about prediction error. To evaluate how exact and precise
is my model predicting concentration, I can apply the error
theory...Calculating relative error (evaluate how exact it is) and
square root of cuadratic error (evaluate how precise it is)?
And relative error is 1/N.(SUM(Cpred-Cpexp)/Cpexp)
And RECM is SQR(SUM(Cpred-Cpexp)2)
Am I right?
Thanks for your help
Joaquim
Back to the Top
Joaquim,
The bootstrap issue has been discussed on nmusers several times. See http://www.cognigencorp.com/nonmem/nm/99jul292006.html
for details and these references:
1. An investigation reported by Mark Gastonguay and Ahmed El-Tahtawy
"Minimization status had minimal impact on the resulting BS
[bootstrap] parameter distributions" http://metrumrg.com/publications/Gastonguay.BSMin.ASCPT2005.pdf
2. An investigation of 13 data sets reported by Nick Holford with Carl
Kirkpatrick and Steve Duffull "NONMEM Termination Status is Not an
Important Indicator of the Quality of Bootstrap Parameter Estimates" http://www.page-meeting.org/default.asp?abstract=992
See this paper for your questions about bias and precision:
Sheiner LB, Beal SL. Some Suggestions for Measuring Predictive
Performance. J Pharmacokinet Biopharm. 1981;9(4):503-12.
Best wishes,
Nick
Want to post a follow-up message on this topic?
If this link does not work with your browser send a follow-up message to PharmPK@boomer.org with "Nonmem bootstrap-prediction error" as the subject | Support PharmPK by using the |
Copyright 1995-2011 David W. A. Bourne (david@boomer.org)