Parameterization differences
From: "Piotrovskij, Vladimir" <vpiotrov@janbe.jnj.com>
Subject: Parameterization differences
Date: 2 Apr 1998 02:09:28 -0500
Dear Alison,
I fitted ADVAN2 TRANS1 model to my data using two parameterizations: with V and CL which should give identical results. However, it was not so. The first one:
$PK
K = THETA(1) * EXP(ETA(1))
V = THETA(2) * EXP(ETA(2))
KA = THETA(3) * EXP(ETA(3))
TLAG = 0
IF (TRT.NE.1) TLAG=THETA(4)
ALAG1= TLAG * EXP(ETA(4))
S2 = V
CL = K*V
$EST MAX=9000 NOABORT METHOD=1 INTERACTION
It converged at 733.243 with the following estimates of THETA
TH 1 TH 2 TH 3 TH 4
9.27E-02 2.16E+02 4.57E+00 2.01E-01
The alternative parameterization:
$PK
K = THETA(1) * EXP(ETA(1))
CL = THETA(2) * EXP(ETA(2))
KA = THETA(3) * EXP(ETA(3))
TLAG = 0
IF (TRT.NE.1) TLAG=THETA(4)
ALAG1= TLAG * EXP(ETA(4))
V = CL/K
S2 = V
$EST MAX=9000 NOABORT METHOD=1 INTERACTION
It converged at 747.266, and estimates were
TH 1 TH 2 TH 3 TH 4
9.27E-02 2.00E+01 4.60E+00 2.06E-01
The rest of the NM-TRAN code was the same. What's also strange: the elapsed CPU time for the first version was about 6 min while for the second version more than 6 h.
Am I missing something?
TIA,
Vladimir
----------------------------------------------------------------------
Vladimir Piotrovsky, Ph.D.
Janssen Research Foundation
Clinical Pharmacokinetics
B-2340 Beerse
Belgium
Fax: +32-14-605834
Email: vpiotrov@janbe.jnj.com
vpiotrov@janbelc1.ssw.jnj.com