From: Mark Sale - Next Level Solutions <*mark*>

Date: Mon, 21 Jul 2008 03:01:46 -0700

<= td style="text-align: left;" align="left" valign="top" width="108">= Thanks Leonid,

= I believe what you tell me, and I understand that FOCE doesn't solve the p= roblem with the approximation that FO makes, only reduces it (and possibly = expands the range that the approximation is useful for?). Anyone out = there with insight into what a practical limit is for FOCE and/or if there = are any diagnostics that are helpful when you're close to it? Is it r= eally 0.5 for FO?

Mark

Mark Sale MD

Next Level Solutions, = LLC

www.NextLevelSolns.com

919-846-91= 85

**
**

Date: Mon, 21 Jul 2008 03:01:46 -0700

<= td style="text-align: left;" align="left" valign="top" width="108">= Thanks Leonid,

= I believe what you tell me, and I understand that FOCE doesn't solve the p= roblem with the approximation that FO makes, only reduces it (and possibly = expands the range that the approximation is useful for?). Anyone out = there with insight into what a practical limit is for FOCE and/or if there = are any diagnostics that are helpful when you're close to it? Is it r= eally 0.5 for FO?

Mark

Mark Sale MD

Next Level Solutions, = LLC

www.NextLevelSolns.com

919-846-91= 85

-------- Original Message --------

Subject: Re: [NMusers] algorithm limits

From: Leonid Gibiansky <LGib= iansky ale - Next Level Solutions <mark k="return true;Popup.composeWindow('pcompose.php?sendto=nmusers%40globo= maxnm.com');; return false;" href="mailto:nmusers ="_blank">www.quantpharm.com

e-mail: LGibiansky at quantpharm.com

tel: (301) 767 5566

Mark Sale - Next Level Solutions wrote:

>

> Leon= id,

> This isn't PK, and the model show basically the right shape, an= d the

> data suggest reasonable residual error (the biological marke= r falls from

> a value between 5 and 310000, to somewhere between 0 = and no change from

> baseline, over a course of a couple of hours to= a couple of weeks, then

> recovers somewhere between a 100 hours an= d 9000 hours later.)

> ie., it start at a highly variable level fall = by some highly variable

> fraction, over some variable lenghth of ti= me and recovers somewhere

> between about a week and about a year.> But, within those limits, it appears pretty well behaved.

>>

> Mark Sale MD

> Next Level Solutions, LLC

> www.NextLevelSolns.com <http://www.NextLevelSolns.com>

> 919-846-9185

>

> -------- Original Message ----= ----

> Subject: Re: [NMusers] algorithm limits

> From: Leonid G= ibiansky <LGibiansky com>

> Date: Sat, July 19, 2008 5:36 pm

> To: Mark Sale = - Next Level Solutions <mark evelsolns.com>

> Cc: nmusers<= /b> pcompose.php#Compose" target="_blank" _onclick="return true;Popup.compo= seWindow('pcompose.php?sendto=nmusers%40globomaxnm.com'); return false;" = mce_href="http://email.secureserver.net/pcompose.php#Compose">nmusers<= /b> u really have 10,000 fold differences in, say, volume or

> bioavailab= ility, population model does not make any sense: individual

> paramet= ers have uninformative priors; they are defined by the

> individual> data only, no meaningful predictions can be made for the next patien= t.

> So, if you need data description, you can directly see whether t= he

> method provides you with the correct line, but you cannot count = on

> prediction: they can be anywhere.

>

> For the estim= ation procedure, my understanding is that large OMEGAs

> will

>= discount population model influence on the individual fit, and in this

= > respect, the method will give you the correct answer (individual

&g= t; parameters controlled by the individual data only). This is how you

&= gt; trick nonmem into the individual model fit: assign huge OMEGAs. Whether=

> your true OMEGA value is 50 or 150 is more or less irrelevant: bot= h

> values are huge and do not provide informative priors for the

= > individual

> parameters.

>

> Sometimes you get huge= OMEGAs if there is a strong correlation between

> parameters, so tha= t combination of ETAs is finite while each of them

> individually can= be anywhere. Removal of some random effects can

> help in

> th= is case. Sometimes large OMEGAs are indicative of multivariate

> dist= ributions (or strong categorical covariate effects): this will be

> s= een on ETA distributions histograms or ETAs vs covariates plots.

>> Overall, I think you have problems with the model or data rather tha= n

> with the estimation method failure.

>

> Thanks

&g= t; Leonid

>

> --------------------------------------

> L= eonid Gibiansky, Ph.D.

> President, QuantPharm LLC

> web:www.quantpharm.com <http://www.quantpharm.com>

> e-mail: LGibians= ky at quantpharm.com <http://quantpharm.com>

>= ; tel: (301) 767 5566

>

>

>

>

> Mark Sale= - Next Level Solutions wrote:

> >

> > General question:<= br>> > What are practical limits on the magnitude of OMEGA that is

> compatible

> > with the FO and FOCE/I method? I seem to reca= ll Stuart at one time

> > suggesting that a CV of 0.5 (exponential= OMEGA of 0.5) was about the

> > limit at which the Taylor expansi= on can be considered a reasonable

> > approximation of the real di= stribution. What about FOCE-I?

> > I'm asking because I have a mod= el that has an OMEGA of 13,

> exponential

> > (and sometime = 100) FOCE-I, and it seems to be very poorly behaved in

> > spite o= f overall, reasoable looking data (i.e., the structural model

> > = traces a line that looks like the data, but some people are WAY

> abo= ve

> > the line and some are WAY below, and some rise MUCH faster,= and some

> > rise MUCH later, by way I mean >10,000 fold, but = residual error

> looks

> > not too bad). Looking at the raw = data, I believe that the the

> > variability is at least this larg= e. Can I beleive that NONMEM FOCE

> > (FO?) will behave reasonably= ?

> > thanks

> > Mark

> >

>

(image/png attachment: left.letterhead)

(image/png attachment: 02-left.letterhead)