```From: clement clement@ha.mc.ntu.edu.tw
Date: Tue, April 27, 2004 9:24 pm

Dear NONMEM users:

Regarding the unit of Km in ADVAN6 and ADVAN10, quite a few discussions
have been posted in NONMEM Users Network. Although so far our lab have
the lucky experience to have a successful case by using both
subroutines, I have been contemplating this old question for some time,
and I wonder if any one could kindly help me to further clarify the
following questions.

First, in ADVAN6, it is recommended to define C2=A(2)/S2 along with
dA/dt = -C2*Vm/(Km+C2) to have Km unit of concentration. One possible
reason has been proposed in 1995 as the following:

" ....But within the \$DES block, unscaled compartment amounts are
present in the variables A(1), A(2), A(3), etc.  This is why A(2)/S2 (a
concentration) was computed for use in the M-M elimination term
-C2*VM/(KM+C2)... "

It is no problem to use "amount" and "time" for R (or dA/dt) because
those are what we input in the data file. However, what we input for the
DV (dependent variable) are usually in the unit of "concentration". Does
"NONMEM" internally multiply the DV (in a unit of concentration) by
volume to have a unit of amount in A(2) so that we have to re-scaled it
again by defining A(2)/S2 in \$DES block (as well as it is used as
F=A(2)/S2 in \$ERROR)? Or, just like what has been told by Dr. Joachim
Grevel:
"    Since she defines the CENTRAL compartment as DEFOBS, the line
>   C2      =  A(2)/S2
in \$DES is redundant ..."

Likewise, in ADVAN10, it is known that ADVAN10 uses amounts rather than
concentrations in its computations. Regardless of its physiological
meaning, again, does "NONMEM" internally multiply the DV (in a unit of
concentration) by volume to have a unit of amount in A(2) to be used in
-A(2)*VM/(KM+A(2))? If so, where does "NONMEM" find the value of volume
to be multiplied?

Any help would be much appreciated!

CJ Lin,
College of Medicine, NTU

_______________________________________________________

```