Hi
Training material distributed during an ICON training contains the following
verbatim statements about mu referencing
"
Code already defined as Typical Value (TV), actual value (as recommended by
Beal) are easy to convert:
– TVCL = THETA(1)*AGE**THETA(2)
– MU_5 = LOG(TVCL)
– CL = EXP(MU_5+ETA(5))
Even better, linear relationship of all THETAS with MU’s:
– LTVCL = THETA(1) + THETA(2)*LOG(AGE)
– MU_5 = LTVCL
– CL = EXP(MU_5+ETA(5))
"
Can someone comment on why the second parameterization is "even better ?
(besides the fact that estimates don't need to be bound)
Thanks
Sebastien Bihorel
********************************************************************
This e-mail and any attachment hereto, is intended only for use by the
addressee(s) named above and may contain legally privileged and/or confidential
information. If you are not the intended recipient of this e-mail, any
dissemination, distribution or copying of this email, or any attachment hereto,
is strictly prohibited. If you receive this email in error please immediately
notify me by return electronic mail and permanently delete this email and any
attachment hereto, any copy of this e-mail and of any such attachment, and any
printout thereof. Finally, please note that only authorized representatives of
Regeneron Pharmaceuticals, Inc. have the power and authority to enter into
business dealings with any third party.
********************************************************************
Clarification on performance of mu referencing
Hi Sébastien,
My understanding is that keeping all thetas additive makes mu r referencing
more efficient. The underlying mu reference math is then accurate for
matrix linear formulations.
My understanding is not perfect for this, so I'll defer to others if there
is a different answer.
Thanks,
Bill
Quoted reply history
On Wed, Aug 23, 2023, 8:11 AM Sébastien Bihorel <
[email protected]> wrote:
> Hi
>
> Training material distributed during an ICON training contains the
> following verbatim statements about mu referencing
>
> "
> Code already defined as Typical Value (TV), actual value (as recommended
> by Beal) are easy to convert:
> – TVCL = THETA(1)*AGE**THETA(2)
> – MU_5 = LOG(TVCL)
> – CL = EXP(MU_5+ETA(5))
>
> Even better, linear relationship of all THETAS with MU’s:
> – LTVCL = THETA(1) + THETA(2)*LOG(AGE)
> – MU_5 = LTVCL
> – CL = EXP(MU_5+ETA(5))
> "
>
> Can someone comment on why the second parameterization is "even better ?
> (besides the fact that estimates don't need to be bound)
>
> Thanks
>
> Sebastien Bihorel
> ********************************************************************
> This e-mail and any attachment hereto, is intended only for use by the
> addressee(s) named above and may contain legally privileged and/or
> confidential information. If you are not the intended recipient of this
> e-mail, any dissemination, distribution or copying of this email, or any
> attachment hereto, is strictly prohibited. If you receive this email in
> error please immediately notify me by return electronic mail and
> permanently delete this email and any attachment hereto, any copy of this
> e-mail and of any such attachment, and any printout thereof. Finally,
> please note that only authorized representatives of Regeneron
> Pharmaceuticals, Inc. have the power and authority to enter into business
> dealings with any third party.
> ********************************************************************
>