RE: negative time values

Andreas: Presently and historically, NONMEM (actually PREDPP) has a restriction against negative time. Removing this restriction by commenting out a piece of code in PREDPP and running a test problem indicates that the differential equation solvers are set up in NONMEM to be able to handle this properly, so there may be no need for the negative time restriction. Further tests need to be made to ensure that the PREDPP routines can handle properly multiple doses given in negative time, etc., and a future version of NONMEM will have the negative time restriction removed. Until then, there is a simple work-around for the present NONMEM as well as previous versions. The following code example from a modification of CONTROL7 demonstrates the use of taking the original TIMEO data item (which is the original time, with negative times), adding a shift to it (the time of the first record in each individual), and generating a new TIME data item that PREDPP has no problem using, as these new times are all non-negative. The data file itself does not need to be altered. The data file does not need an extra column for the altered TIME, but just contains the original time column TIMEO, as identified in the $INPUT record. The $INPUT record also ends with a TIME item for the 7th column, yet there are only 6 columns in the data set. NMTRAN will complain that there are more $INPUT items than there are columns in the data file, so the $WARN DATA=1 is set so only a single warning shows up (just to remind you). NMTRAN will fill in this non-existent seventh column with null in FDATA, which NONMEM then uses. This technique creates a place holder for the altered time that $INFN fills in at run-time. Therefore, both TIME and TIMEO can be printed in a table. By the way, start_time is easily retrieved as STIM=TIMEO-TIME, also available for table output. This technique of using $INFN during ICALL=-1, that is before the estimation begins, is useful for on-the-fly modification of the input data by NONMEM, and can be used for other purposes. $PROB THEOPHYLLINE POPULATION DATA $WARN DATA=1 ; suppress warnings about missing TIME data item in data file $INPUT ID DOSE=AMT TIMEO CP=DV WT EVID MDV TIME ; the data file contains $DATA THEOPPMINUS.csv $SUBROUTINES ADVAN6 TOL=5 $MODEL COMP=(DEPOT,INITIALOFF,DEFDOSE) COMP=(CENTRAL,DEFOBS,NOOFF) $INFN IF (ICALL==1) THEN DOWHILE(DATA) IF (NEWIND<2) START_TIME=TIMEO TIME=TIMEO-START_TIME ENDDO ENDIF $PK ;THETA(1)=MEAN ABSORPTION RATE CONSTANT (1/HR) ;THETA(2)=MEAN ELIMINATION RATE CONSTANT (1/HR) ;THETA(3)=SLOPE OF CLEARANCE VS WEIGHT RELATIONSHIP (LITERS/HR/KG) ;SCALING PARAMETER=VOLUME/WT SINCE DOSE IS WEIGHT-ADJUSTED KA=THETA(1)+ETA(1) KE=THETA(2)+ETA(2) CL=THETA(3)*WT+ETA(3) S2=CL/KE/WT STIM=TIMEO-TIME $THETA (.1,3,5) (.008,.08,.5) (.004,.04,.9) $OMEGA BLOCK(3) 6 .005 .0002 .3 .006 .4 $DES DADT(1)=-KA*A(1) DADT(2)= KA*A(1)-KE*A(2) $ERROR Y=F+EPS(1) $SIGMA .4 $EST MAXEVAL=450 PRINT=5 $COV $TABLE ID DOSE WT STIM TIMEO TIME NOPRINT FILE=CONTROL7MINUS.TAB $SCAT (RES WRES) VS TIME BY ID The following is part if the data file. Note that the altered TIME column (column 7) is not necessary to be present 1,4.02,-2,.,79.6,1,1 1,.,-2,0.74,.,0,0 1,.,-1.75,2.84,.,0,0 1,.,-1.43,6.57,.,0,0 1,.,-0.88,10.5,.,0,0 1,.,0.02,9.66,.,0,0 1,.,1.82,8.58,.,0,0 1,.,3.1,8.36,.,0,0 1,.,5.03,7.47,.,0,0 1,.,7.05,6.89,.,0,0 1,.,10.12,5.94,.,0,0 1,.,22.37,3.28,.,0,0 2,4.4,-2,.,72.4,1,1 2,.,-2,0,.,0,0 2,.,-1.73,1.72,.,0,0 2,.,-1.48,7.91,.,0,0 2,.,-1,8.31,.,0,0 2,.,-0.08,8.33,.,0,0 2,.,1.5,6.85,.,0,0 2,.,3.02,6.08,.,0,0 2,.,5.03,5.4,.,0,0 2,.,7,4.55,.,0,0 2,.,10,3.01,.,0,0 2,.,22.3,0.9,.,0,0 Robert J. Bauer, Ph.D. Vice President, Pharmacometrics, R&D ICON Development Solutions 7740 Milestone Parkway Suite 150 Hanover, MD 21076 Tel: (215) 616-6428 Mob: (925) 286-0769 Email: [email protected]<mailto:[email protected]> Web: http://www.iconplc.com/