- Contemporary messages sorted: [ by date ] [ by thread ] [ by subject ] [ by author ] [ by messages with attachments ]

From: Jeroen Elassaiss-Schaap (PD-value B.V.) <"Jeroen>

Date: Tue, 22 Dec 2015 10:26:15 +0100

Dear Anyue,

It has been a long time since I used this method. Nowadays NONMEM has

much added functionality, such as the non-parameteric step and the

possibility to repeat problems over individuals in your dataset. My

advise would actually be to use those. The latter especially gives you a

much cleaner implementation of extended least squares if that is what

you are looking for.

Anyway, your code seems fine. Your entries into $THETA and $SIGMA

however do matter. I would suggest to run this once, and than take the

median of parameters and input that into $THETA. Also take the variance

of residuals and put that number into $SIGMA. Now run it again, and the

result should be unbiased. I do not have a reference for this recipe,

please use this at your own discretion.

Hope this helps,

Jeroen

http://pd-value.com

jeroen_at_pd-value.com

_at_PD_value

+31 6 23118438

-- More value out of your data!

On 12/22/2015 07:36 AM, Anyue Yin wrote:

*> Dear all,
*

*>
*

*> I got a problem when using NONMEM with STS method. My aim is to get
*

*> individual parameter estimates. I would like to let NONMEM estimate
*

*> individual parameter one subject by one subject. My control stream is
*

*> partially listed below, which is edited according to this thread
*

*> http://www.cognigencorp.com/nonmem/nmo/topic035.html
*

*>
*

*> /$SUBROUTINES ADVAN1 TRANS2/
*

*> /
*

*> /
*

*> /$PK/
*

*> /CL = THETA(1) * EXP(ETA(1))/
*

*> /V = THETA(2) * EXP(ETA(2))/
*

*> /S1 = V/
*

*> /
*

*> /
*

*> /$ERROR/
*

*> /IPRED = F/
*

*> / W = F/
*

*> / Y = IPRED*(1+EXP(ETA(3))*EPS(1))/
*

*> / IRES = DV-IPRED/
*

*> /IWRES = IRES/W/
*

*> /
*

*> /
*

*> /$THETA/
*

*> /(10,20,30) ; CL/
*

*> /(10,80,100) ; V/
*

*> /
*

*> /
*

*> /$OMEGA/
*

*> /100 ; IIV CL/
*

*> /100 ; IIV V/
*

*> /100 ; IIV SIGMA/
*

*> /
*

*> /
*

*> /$SIGMA/
*

*> /1 FIXED ; PRO/
*

*> /$EST METHOD=1 INTER MAXEVAL=0 NOABORT SIG=3 PRINT=1 POSTHOC/
*

*> /$COV/
*

*> /$TABLE ID TIME AMT CMT DV PRED IPRED IWRES CL V ETA(1) ETA(2) ETA(3)
*

*> ONEHEADER NOPRINT FILE=sdtab101.tab/
*

*>
*

*> The model can run successfully, and I got individual parameter
*

*> estimates from output file. I found that for each subject the
*

*> individual parameter estimate changed if I change the initial value
*

*> of THETA. It seems that NONMEM regarded the initial value fixed (just
*

*> like a fixed typical value). But I expected NONMEM to estimate
*

*> individual parameters based on individual data rather than the initial
*

*> value. Is there anything wrong with the code? Thanks in advance.
*

*>
*

*> Anyue
*

Received on Tue Dec 22 2015 - 04:26:15 EST

Date: Tue, 22 Dec 2015 10:26:15 +0100

Dear Anyue,

It has been a long time since I used this method. Nowadays NONMEM has

much added functionality, such as the non-parameteric step and the

possibility to repeat problems over individuals in your dataset. My

advise would actually be to use those. The latter especially gives you a

much cleaner implementation of extended least squares if that is what

you are looking for.

Anyway, your code seems fine. Your entries into $THETA and $SIGMA

however do matter. I would suggest to run this once, and than take the

median of parameters and input that into $THETA. Also take the variance

of residuals and put that number into $SIGMA. Now run it again, and the

result should be unbiased. I do not have a reference for this recipe,

please use this at your own discretion.

Hope this helps,

Jeroen

http://pd-value.com

jeroen_at_pd-value.com

_at_PD_value

+31 6 23118438

-- More value out of your data!

On 12/22/2015 07:36 AM, Anyue Yin wrote:

Received on Tue Dec 22 2015 - 04:26:15 EST