simulation with nonparametric method in NONMEM

From: nele.kaessner
Date: Fri, 3 Jul 2009 12:53:50 +0200

Dear all,

I have a question concerning the nonparametric estimation in NONMEM. I
used the nonparametric method for phase I single dose PK data because
parameters seemed not to be normally distributed. The dataset contained 40
subjects and each received 3 different doses. The model fit the data well,
and as a last evaluation I wanted to simulate 500 new subjects for each
dose level and compare the simulations to the observed values (in a
spaghetti plot which includes all simulated profiles as half-transparent,
which enables the identification of concentrations which are repeated more
often than others, as the respective area is darker). For comparison, I
wanted to use the IPRED values and disregard the residual error component.
What puzzled me when I saw the plot was that I did only see as many lines
as I had subjects in the trial. So my question is: If I only have a
limited number of subjects, does it make sense to use a nonparametric
method at all, because the simulations I get from this seem to be a bit
limited? Do I only get the exact parameter combination as I have already
observed with my subjects, or can a nonparametric method also simulate
other parameter combinations (in my case I have IIV on three different
parameters)? The used model code is shown below.

Thanks and best regards
Nele

\$MODEL COMP=(GUT) COMP=(CENTRAL) COMP=(PERIPH) COMP=(BINDING)
;
\$PK

JD=DEN_
DN1=CDEN_(1)
DN2=CDEN_(2)
DN3=CDEN_(3)

;
TVCL=THETA(1)
CL=TVCL*EXP(ETA(1))

TVV2=THETA(2)
V2=TVV2*EXP(ETA(3))

TVQ=THETA(3)
Q=TVQ

TVV3=THETA(4)
V3=TVV3

TVKA=THETA(5)
KA=TVKA

TVF1=THETA(6)
F1=TVF1*EXP(ETA(2))

TVLAG=THETA(7)
ALAG1=TVLAG

TVK1=THETA(8)
K1=TVK1

TVK2=THETA(9)
K2=TVK2

TVBMAX=THETA(10)
BMAX=TVBMAX

S2=V2
K23=Q/V2
K32=Q/V3
K20=CL/V2

\$DES

;
\$ERROR
IPRED=F
DEL=0
IF (IPRED.EQ.0) DEL=0.0001
W=F
IRES=DV-IPRED
IWRES=IRES/(W+DEL)
Y=F+SQRT(THETA(12)*THETA(12)+THETA(11)*THETA(11)*F**2)*EPS(1)

\$MSFI=msfb1
\$SIMULATION (12245) (123456 NONPARAMETRIC) ONLYSIM TRUE=FINAL
SUBPROBLEMS=500
______________________________________________________________

Dr. Nele Käßner
Pharmacometrics -- Modeling and Simulation

Nycomed GmbH
Byk-Gulden-Str. 2
D-78467 Konstanz, Germany

Fon: (+49) 7531 / 84 - 4759
Fax: (+49) 7531 / 84 - 94759

mailto: nele.kaessner
http://www.nycomed.com

County Court: Freiburg, Commercial Register HRB 701257
Chairman Supervisory Board: Charles Depasse
Management Board: Dr. Barthold Piening, Gilbert Rademacher, Dr. Anders
Ullman

----------------------------------------------------------------------
Proprietary or confidential information belonging to Nycomed Group may
be contained in this message. If you are not the addressee indicated
in this message, please do not copy or deliver this message to anyone.
In such case, please destroy this message and notify the sender by