[NMusers] PK/PD model with a categorical variable

From: SIMON Nicolas <Nicolas.SIMON_at_ap-hm.fr>
Date: Sun, 25 Jun 2017 20:37:26 +0000

Dear NMusers,


I tried to fit a PK/PD model with a categorical variable (0/1).


The PK part was estimated in a previous run (2 steps process with fixed the=
ta, omega and sigma for PK).


The hypothesis for the categorical variable is a value of 1 with high conce=
ntrations and 0 otherwise.


Something is wrong in the $ERROR as I observed that the F values used for c=
omputing EXPP were not realistic (far to high).


Any advice will be much appreciate.

Best regards

Nicolas



$PROB PK/PD
$INPUT ID TIME DAT1=DROP DV AMT EVID MDV TYPE MTT PERIOD DOSE GROUPE RATE
$DATA ..\data.csv
 IGNORE=#
$ESTIM MAXEVAL=9000 SIGDIG=2 PRINT=10 NOABORT METHOD=CONDITIONAL IN=
TERACTION LAPLACIAN
$COV
$THETA (440 FIXED) ;POPV1 ;L
$THETA (0.852 FIXED) ;POPK10 ;1/H
$THETA (2.13 FIXED) ;POPK12 ;1/H
$THETA (0.069 FIXED) ;POPK21 ;1/H
$THETA (3.98 FIXED) ;POPK13 ;1/H
$THETA (1.67 FIXED) ;POPK31 ;1/H
$THETA (2.41 FIXED) ;COV_GR
$THETA (0.677 FIXED) ;COV_DO
$THETA (.1) ;int
$THETA (.1) ;slp
$OMEGA 0 FIXED ; ETA_V1
$OMEGA 1.12 FIXED ; ETA_K10
$OMEGA 0 FIXED ; ETA_K12
$OMEGA 0 FIXED ; ETA_K21
$OMEGA 0 FIXED ; ETA_K13
$OMEGA 0 FIXED ; ETA_K31
$OMEGA 0.717 FIXED ; ETA_F1
$SIGMA 0.126 FIXED ; ERR1

$SUBR ADVAN11 TRANS1
$PK
   D1=0.1667
   GRP=0
   IF(GROUPE.EQ.2) GRP=1
   GRP_DO=0
   IF(DOSE.GT.15) GRP_DO=1
   TVF1=THETA(7)**GRP * THETA(8)**GRP_DO
   F1=TVF1*EXP(ETA(7))
   V=THETA(1)*V1*EXP(ETA(1))
   K=THETA(2)*EXP(ETA(2))

   K12=THETA(3)*EXP(ETA(3))

   K21=THETA(4)*EXP(ETA(4))
   K13=THETA(5)*EXP(ETA(5))
   K31=THETA(6)*EXP(ETA(6))
   S1=V
   CL=K*V
   INT=THETA(9)
   SLP=THETA(10)

$ERROR
  DEL=0
  IF (F.EQ.0) DEL=0.0001
  W=F+DEL
  EXPP=INT+F*SLP
  IF(EXPP.GT.30.0) EXPP=30.0

IF (TYPE.EQ.0) THEN
    F_FLAG=0
    Y=F + W*ERR(1)
ENDIF
IF (TYPE.EQ.1) THEN
    F_FLAG=1
    AA=DEXP(EXPP)
    BB=1+AA
    Y=DV*AA/BB+(1-DV)/BB
ENDIF


$TABLE ID TIME AMT DOSE GROUPE DV CWRES PRED EVID MDV TYPE

FILE=r



#ID TIME DATE DV AMT EVID MDV TYPE PERIOD DOSETH GROUPE RATE 1 10:04:00 07/=
02/2014 0 89.9984099 1 1 0 1 30 2 -2 1 10:04:01 07/02/2014 0.00248052 0 0 0=
 0 1 30 2 0 1 10:04:01 07/02/2014 0 0 0 0 1 1 30 2 0 1 10:30:00 07/02/2014 =
0.05406265 0 0 0 0 1 30 2 0 1 10:30:00 07/02/2014 1 0 0 0 1 1 30 2 0 1 10:4=
5:00 07/02/2014 0.04515821 0 0 0 0 1 30 2 0 1 10:45:00 07/02/2014 0 0 0 0 1=
 1 30 2 0 1 11:50:00 07/02/2014 0.03370965 0 0 0 0 1 30 2 0 1 11:50:00 07/0=
2/2014 0 0 0 0 1 1 30 2 0 1 12:25:00 07/02/2014 0.01590078 0 0 0 0 1 30 2 0=
 1 12:25:00 07/02/2014 0 0 0 0 1 1 30 2 0 1 14:27:00 07/02/2014 0.00477023 =
0 0 0 0 1 30 2 0 1 14:27:00 07/02/2014 0 0 0 0 1 1 30 2 0 1 16:31:00 07/02/=
2014 0.00314835 0 0 0 0 1 30 2 0 1 16:31:00 07/02/2014 0 0 0 0 1 1 30 2 0 1=
 18:31:00 07/02/2014 0.00298935 0 0 0 0 1 30 2 0 1 20:36:00 07/02/2014 0.00=
219431 0 0 0 0 1 30 2 0 1 22:30:00 07/02/2014 0.0020353 0 0 0 0 1 30 2 0 1 =
10:25:00 08/02/2014 0.00162188 0 0 0 0 1 30 2 0 1 10:25:00 08/02/2014 0 0 0=
 0 1 1 30 2 0 1 09:49:00 21/02/2014 0 26.3952934 4 1 0 2 10 2 -2 1 09:49:01=
 21/02/2014 0.00159008 0 0 0 0 2 10 2 0 1 09:49:01 21/02/2014 0 0 0 0 1 2 1=
0 2 0




Received on Sun Jun 25 2017 - 16:37:26 EDT

This archive was generated by hypermail 2.3.0 : Fri Sep 27 2019 - 16:55:30 EDT