NONMEM requires user specify the statistical structure and the nonlinear dynamics of a model. general, a subroutine ¹ called PRED (stands for "PREDiction"), is required from the user to specifying the likelihood structure based on data and model parameters. However, the PRED can be very complicated (especially when ordinary differential equations are involved) and writing it could be demanding for scientists. A library of PRED routines called PREDPP (stands for PRED Population Pharmacokinetics) has been created for the population pharmacokinetics analysis. As one will see, it is trivial to apply PREDPP to single subject data.

PREDPP contains a number of basic kinetic models for user to select. For example: if the user is interested in a one compartment linear model, he chooses subroutine from a list of PREDPP Library subroutines: ADVAN1, ADVAN2, ADVAN3, etc. ²

NM-TRAN stands for "NONMEM Translator". It is a stand-alone program that preprocess input control stream and data, so that they can be used by NONMEM for model fitting. In NONMEM applications almost all model specification is done through NM-TRAN, including all the examples in this documentation. It converts them into three things: a NONMEM data set, a NONMEM control stream ¹, and needed subroutines.

NM-TRAN also supports PREDPP by generating its subroutines. It also handles additional model inputs such mixture models (MIX) and Bayesian inference (PRIOR and BAYES). Unless specified, hereon in this document by "PREDPP reocords/entries" we always refer to when they are used in NM-TRAN.

NM-TRAN takes its input from two files:

• NM-TRAN control stream,
• NM-TRAN data set.

The first file uses control record entry $DATA to specify the path to the second.

NM-TRAN modifies (see next section) its data file input and generates the data file FDATA for NONMEM ². This FDATA file is among the several NM-TRAN output files, including

See reference manual for additional files and details.

NONMEM consumes the input files created by NM-TRAN, and generates a number of files during the run.

If an Model Specification File (MSF) is output, and intermediate output with iteration summaries is requested (i.e., $ESTIM options MSFI and PRINT are used), then NONMEM also writes the parameter estimates of these summaries to the console and to file INTER, which now exists after the run terminates. INTER now include lines giving the parameter estimates in their natural unscaled space (NPARAMETR) as well as the UCP values (PARAMETER).

Additional output files include files with extensions (suffixes) .ext, .cov, .coi, .cor, .phi, .phm, .grd, .shk, .shm, .cnv, smt, .rmt, etc. The names of the files start with "root", where root is the root name of the NM-TRAN control stream file. The file root.ext is called the raw output file. The root name may be specified by $ESTIMATION record option FILE.