Behavioural/Result differences for NONMEM version 5.

Hi, I know we all predominantly use NONMEM versions after NONMEM version 5. But we retain old versions to reproduce or work on historic activities. Anyway the question/issue: We are upgrading our infrastructure from RedHat 4 to RedHat 5 and Intel compiler from 10 to 12. We now have models that worked historically but we get anomalous behaviour varying from differences in theta and eta estimates in the 3rd significant digit up to runs that hang during the convergence steps. Has anyone else experienced these? Note: we have been informed of two compiler warnings during the installation phase: ifort: command line remark #10010: option '-Vaxlib' is deprecated and will be removed in a future release. See '-help deprecated' ifort: command line remark #10279: option '-xW' is deprecated and will be removed in a future release. See '-help deprecated' Your feedback/advice on this is most welcome. Regards, Chris Franklin ________________________________ This e-mail was sent by GlaxoSmithKline Services Unlimited (registered in England and Wales No. 1047315), which is a member of the GlaxoSmithKline group of companies. The registered address of GlaxoSmithKline Services Unlimited is 980 Great West Road, Brentford, Middlesex TW8 9GS.

Hi Chris, It could be that the SLOW option of the $EST step may be of use here (see the notes in nmhelp) - did you try that / does that help? Erik

Dear Chris, You are switching both NONMEM versions and compiler versions and operating system versions. I'd be amazed if you found identical results on all your models. Only the very simple or very well defined ones will not move around, but for any real-life model, numerical outcome will depend on all these factors plus additional influences like the operating system itself (i.e. Linux, Windows, OSX), running 32 or 64 bit versions of OS and/or software, the whole multitude of compiler switches, running NONMEM with or without MPI, and finally hardware differences like Intel or AMD processors. To get back to your question: I'd be surprised if the Redhat version would affect your results but changing the Intel compiler version almost certainly will (unless you've been particularly fortunate with your choice of compiler switches) and so will changing NONMEM version. Good luck, Rik Schoemaker, PhD Life Science Services - Exprimo Senior Consultant Phone: +31 20 441 6410 E-mail: [email protected] <mailto:[email protected]%20> Web: http://www.sgs.com/exprimo SGS Exprimo NV Generaal De Wittelaan 19A Bus 5 B-2800 Mechelen, Belgium Switchboard: +32 15 27 32 45 This e-mail is confidential. It is also privileged or otherwise protected by work product immunity or other legal rules. The information is intended to be for use of the individual or entity named above. If you are not the intended recipient, please be aware that any disclosure, copying, distribution or use of the contents of this information is prohibited. You should therefore delete this message from your computer system. If you have received the message in error, please notify us by reply e-mail. The integrity and security of this message cannot be guaranteed on the Internet. Thank you for your co-operation.