From: YZ [mailto:zhang.239@osu.edu] Subject: [NMusers] How to do simulation? Date: Thursday, November 06, 2003 1:57 PM Dear All, I am working on an oral compound that can be described by one compartment model with 1st order absorption and 1st order elimination. Preclinical PK results show that most of the absorbed dose was metabolized to one metabolite, and unchanged drug was excreted in urine, bile and feces. The clinical results show that the plasma concentration of the metabolite is proportional to its parent compound. But the AUC ratio between the metabolite and parent compound was variant with time (different day). I built a model to describe this phenomenon. My question is how to do the simulation to the metabolite in NONMEM? I can get an output table with DV and MET and also PRED for the DV, but I don't know how to output the predicted value for the metabolite. Thanks. Yilong Zhang, Ph.D. Postdoctoral Researcher College of Pharmacy The Ohio State University _______________________________________________________ From: Bhattaram, Atul [mailto:BhattaramA@cder.fda.gov] Subject: RE: [NMusers] How to do simulation? Date: Thursday, November 06, 2003 2:45 PM Hello Zhang I am not fully clear with your control stream but you can output PRED for both drug and metabolite. If you could send me your control stream I can perhaps help you. Venkatesh Atul Bhattaram CDER, FDA. _______________________________________________________ From: "Bachman, William (MYD)" bachmanw@iconus.com Subject: RE: [NMusers] How to do simulation? Date: 11/6/2003 3:08 PM Write the compartmental model that represent the plausible biological model including the parent and metabolite (using general linear models or differential equation defined model). Create a data file containing the dose records, the observation records for the parent and observation records for the metabolite (in chronological order of course). Include the CMT and EVID data items. Dose records will be CMT=# of compartment where dosed and EVID=0, parent observation records will have CMT=# of compartment where parent is observed and EVID=1, and the metabolite records will have CMT=# of compartment where metabolite is observed and EVID=1. If there are times that you would like predictions that you don't have observations, insert records with the appropriate CMT (for parent or metabolite) and EVID=2. The prediction is always the prediction for the appropriate CMT if it is coded, otherwise the PRED defaults are used. Is this what you are asking? _______________________________________________________ From: "YZ"Subject: RE: [NMusers] How to do simulation? Date: Thu, November 6, 2003 4:15 pm Dear All: Following is the model I wrote. Please let me know how to change it to get the simulation for both parent compound and metabolite. Thanks a lot. Yilong $PROB ALL_MET $INPUT ID AMT TIME DV MET $DATA ..\all_met.csv $SUBROUTINES ADVAN6 TOL=3 $MODEL COMP=(DEPOT,INITIALOFF,DEFDOSE) COMP=(CENTRAL,DEFOBS,NOOFF) COMP=(EFFECT) $PK CALLFL=1 KA=THETA(1)*(1+ETA(1)) KE=THETA(2)*(1+ETA(2)) KM=THETA(3)*(1+ETA(3)) KEM=THETA(4)*(1+ETA(4)) CL=THETA(5)*(1+ETA(5)) S2=CL/(KE+KM) $THETA(0,1)(0,0.04)(0,0.04)(0,0.03)(0,10) $OMEGA 0.5 0.5 0.5 0.5 0.5 $DES DADT(1)=-KA*A(1) DADT(2)=KA*A(1)-(KE+KM)*A(2) DADT(3)=KM*A(2)-KEM*MET*(CL/(KE+KM)) $ERROR Y=F*(1+EPS(1)) $SIGMA 0.5 $ESTIMATION MAXEVAL=999 PRINT=5 POSTHOC $COVARIANCE $SCATTER TIME VS DV $SCATTER TIME VS PRED $SCATTER TIME VS RES $SCATTER PRED VS DV UNIT _______________________________________________________