RE: Large Numbers of Compartments
Dear William,
Three suggestions:
1) If you simulate out 5 vs. 10 vs. 20 transit compartments, the difference
in how sharp a signal switches between 5 and 95% may or may not make
a meaningful difference for your system.
So, it may be worthwhile to limit the no. of transit compartments
to 5 per process and this might get you below 100 comp. in total.
2) I have seen three colleagues using ADAPT5 (or ADAPT II rel. 4?) with a couple
of hundred compartments using a pooled fit approach, as far as I remember.
I have never reached this limit, however, am reasonably sure that ADAPT5
and S-ADAPT will handle such large systems with >100 differential equations
in a nonlinear mixed-effects mode.
David D'Argenio once told me that ADAPT has been built also to solve engineering
problems of hundreds of diff. equations. The default number is S-ADAPT 1.56 is
120 compartments. This number can be increased in the globals.inc.
3) Your system is an ideal one to be parallelized, since solving those
diff. eqs. will take a lot of time. The next release of NONMEM will
do this for you. S-ADAPT has it implemented since a couple of years.
Please let me know, if you like to try S-ADAPT.
Best wishes
Juergen
Quoted reply history
From: [email protected] [mailto:[email protected]] On
Behalf Of Denney, William S.
Sent: Thursday, August 05, 2010 3:41 PM
To: [email protected]
Subject: [NMusers] Large Numbers of Compartments
Hello,
I have a model where I will likely need between 130 and 150 compartments (it's
a rich data set and many transit compartments, so it is probably estimable).
When I was looking in SIZES, it indicated that the maximum number of
compartments is 99; I was wondering if there is a way around this limitation?
Alternatively, is there another tool that people would suggest that can handle
mixed effects with a large number of diff eqs and dosing?
Thanks,
Bill
Notice: This e-mail message, together with any attachments, contains
information of Merck & Co., Inc. (One Merck Drive, Whitehouse Station,
New Jersey, USA 08889), and/or its affiliates Direct contact information
for affiliates is available at
http://www.merck.com/contact/contacts.html) that may be confidential,
proprietary copyrighted and/or legally privileged. It is intended solely
for the use of the individual or entity named on this message. If you are
not the intended recipient, and have received this message in error,
please notify us immediately by reply e-mail and then delete it from
your system.