Hi, NM users,
I have a simple question about how TIME works for multiple doses in NONMEM.
Here is my model:
1-CM with dual absorption process (based on a LAG variable)
IOV is also modeled.
Dose regimen: TID doses at 400 mg
Here is the part of dataset:
C ID TIME DOSE DV CMT EVID MDV AMT WT
1 0 400 . 1 1 1 400 70
1 0 400
2 0 0
70
1 0.25 400 240 2 0 0
70
1 0.5 400 410 2 0 0
70
1 4 400 1420 2 0 0
70
1 6 400 820 2 0 0
70
1 7.9 400 450 2 0 0
70
1 8 400 . 1 1 1 400 70
1 8.25 400 240 2 0 0
70
1 8.5 400 410 2 0 0
70
1 12 400 1420 2 0 0
70
1 14 400 820 2 0 0
70
1 15.6 400 450 2 0 0
70
1 16 400 . 1 1 1 400 70
......................................................
Here is the part of NONMEM control stream:
$PROB PK-TEST
$DATA PK-all.csv IGNORE = C
$INPUT C,ID,TIME,DOSE,DV,CMT,EVID,MDV,AMT,WT
$SUBROUTINE ADVAN2 TRANS2
$PK
OCC1=0
OCC2=0
OCC3=0
OCC4=0
OCC5=0
IF(TIME.LE.12) OCC1=1
IF(TIME.GT.12.AND.TIME.LE.24) OCC2=1
IF(TIME.GT.24.AND.TIME.LE.300) OCC3=1
IOV1=OCC1*ETA(4)+OCC2*ETA(5)+OCC3*ETA(6)
IOV2=OCC1*ETA(9)+OCC2*ETA(10)+OCC3*ETA(11)
TKA1= THETA(3)*DEXP(ETA(1))
TKA2= THETA(5)*DEXP(ETA(2))
LAG=0.75
FLAG=0
IF(TIME.GT.LAG) FLAG=1
TVKA=(1-FLAG)*TKA1 + FLAG*TKA2
........
I highlighted TIME in the code, and my question is that when NONMEM hits LAG,
is TIME reseted? (TIME will start from zero for each dose?) Otherwise, after
0.75 hr, FLAG (even for second dose (0-0.75hr)) will be always equal to 1.
Secondly, during IOV calculation, is TIME relative time to the first dose?
It is really confusing me in the code since the same name, TIME, was used here.
Thanks in advance.
Jian
---------------------------------
Get the Yahoo! toolbar and be alerted to new email wherever you're surfing.
How does TIME work for multiple doses in NONMEM?
TIME is not reset, it is counted from the first dose in this example. To describe the model you need to introduce TAD (time after the most recent dose) into the data set and use it for FLAG reset:
IF(TAD.GT.LAG) FLAG=1
Same with occasion, it would be best to introduce variable OCC (occasion, 1 2
or 3) to the data set
Leonid
Jian Xu wrote:
> Hi, NM users,
>
> I have a simple question about how TIME works for multiple doses in NONMEM.
>
> Here is my model:
> 1-CM with dual absorption process (based on a LAG variable)
> IOV is also modeled.
> Dose regimen: TID doses at 400 mg
>
> Here is the part of dataset:
> Here is the part of NONMEM control stream:
>
> $PROB PK-TEST
> $DATA PK-all.csv IGNORE = C
> $INPUT C,ID,TIME,DOSE,DV,CMT,EVID,MDV,AMT,WT
> $SUBROUTINE ADVAN2 TRANS2
> $PK
> OCC1=0
> OCC2=0
> OCC3=0
> OCC4=0
> OCC5=0
> IF(TIME.LE.12) OCC1=1
> IF(TIME.GT.12.AND.TIME.LE.24) OCC2=1
> IF(TIME.GT.24.AND.TIME.LE.300) OCC3=1
> IOV1=OCC1*ETA(4)+OCC2*ETA(5)+OCC3*ETA(6)
> IOV2=OCC1*ETA(9)+OCC2*ETA(10)+OCC3*ETA(11)
>
> TKA1= THETA(3)*DEXP(ETA(1))
> TKA2= THETA(5)*DEXP(ETA(2))
> LAG=0.75
> FLAG=0
> IF(TIME.GT.LAG) FLAG=1
> TVKA=(1-FLAG)*TKA1 + FLAG*TKA2
> ........
>
> I highlighted TIME in the code, and my question is that when NONMEM hits LAG, is TIME reseted? (TIME will start from zero for each dose?) Otherwise, after 0.75 hr, FLAG (even for second dose (0-0.75hr)) will be always equal to 1. Secondly, during IOV calculation, is TIME relative time to the first dose? It is really confusing me in the code since the same name, TIME, was used here.
>
> Thanks in advance.
>
> Jian
Hi, Leonid,
For some reason, I did not receive your email, and one of my colleagues
forwarded your email to me. Anyway, thanks for your input, and I appreciate
your help.
Actually, when I read that control stream with dataset first time, I had some
kind of feeling that this LAG was not working for multiple dose regimen because
time was not reset. Then after I sent out help to NMusers, I did a
simulation, and output IND and KA values, and it turned out that IND and TVKA
stay the same values after 0.75 hr (TIME in NM was actual relative time.)
Thanks for your suggestion, and I will start to modify the original control
stream.
Have a nice day,
Jian
Leonid Gibiansky <[EMAIL PROTECTED]> wrote: Date: Wed, 06 Jun 2007 12:05:07
-0400
From: Leonid Gibiansky <[EMAIL PROTECTED]>
To: Jian Xu <[EMAIL PROTECTED]>
CC: [email protected]
Subject: Re: [NMusers] How does TIME work for multiple doses in NONMEM?
TIME is not reset, it is counted from the first dose in this example. To
describe the model you need
to introduce TAD (time after the most recent dose) into the data set and use it
for FLAG reset:
IF(TAD.GT.LAG) FLAG=1
Same with occasion, it would be best to introduce variable OCC (occasion, 1 2
or 3) to the data set
Leonid
Jian Xu wrote:
> Hi, NM users,
>
> I have a simple question about how TIME works for multiple doses in NONMEM.
>
> Here is my model:
> 1-CM with dual absorption process (based on a LAG variable)
> IOV is also modeled.
> Dose regimen: TID doses at 400 mg
>
> Here is the part of dataset:
> Here is the part of NONMEM control stream:
>
> $PROB PK-TEST
> $DATA PK-all.csv IGNORE = C
> $INPUT C,ID,TIME,DOSE,DV,CMT,EVID,MDV,AMT,WT
> $SUBROUTINE ADVAN2 TRANS2
> $PK
> OCC1=0
> OCC2=0
> OCC3=0
> OCC4=0
> OCC5=0
> IF(TIME.LE.12) OCC1=1
> IF(TIME.GT.12.AND.TIME.LE.24) OCC2=1
> IF(TIME.GT.24.AND.TIME.LE.300) OCC3=1
> IOV1=OCC1*ETA(4)+OCC2*ETA(5)+OCC3*ETA(6)
> IOV2=OCC1*ETA(9)+OCC2*ETA(10)+OCC3*ETA(11)
>
> TKA1= THETA(3)*DEXP(ETA(1))
> TKA2= THETA(5)*DEXP(ETA(2))
> LAG=0.75
> FLAG=0
> IF(TIME.GT.LAG) FLAG=1
> TVKA=(1-FLAG)*TKA1 + FLAG*TKA2
> ........
>
> I highlighted TIME in the code, and my question is that when NONMEM hits
> LAG, is TIME reseted? (TIME will start from zero for each dose?)
> Otherwise, after 0.75 hr, FLAG (even for second dose (0-0.75hr)) will be
> always equal to 1. Secondly, during IOV calculation, is TIME relative
> time to the first dose?
>
> It is really confusing me in the code since the same name, TIME, was
> used here.
>
> Thanks in advance.
>
> Jian
>
---------------------------------
Building a website is a piece of cake.
Yahoo! Small Business gives you all the tools to get online.
---------------------------------
Boardwalk for $500? In 2007? Ha!
Play Monopoly Here and Now (it's updated for today's economy) at Yahoo! Games.