[Population Modeling] Assumption in models RE: Digest, Vol 12, Issue 5

Tue Oct 6 14:22:37 PDT 2015

Hi John,

You are right, the discussion started from estimation techniques. Yet model
representation is determined by the modeler before any parameter estimation
takes place. In the process of deciding on model representation, the
modeler already made several assumptions, some explicit and some hidden.

Therefore no model is assumption free. Moreover, in my mind a model itself
is an assumption representing how we think reality works.

And just like the scientific process we need to experiment and cross
validate different models to figure out which represent reality better and
under what conditions.

It would be interesting to know how other modelers handle assumptions in
their modeling as well as how those interact with estimation techniques.

I hope more will join this discussion.

                 Jacob

On Fri, Oct 2, 2015 at 9:48 AM, John Rice <john.rice at noboxes.org> wrote:

> Jelliffe
>
> Roger's (Jelliffe) discussions of the basics associated with assumptions
> related statistical analysis and the appropriateness of any particular data
> types or distributions are another indication of needing to systemically
> address all sorts and matters of ASSUMPTION and computational modeling.
> Is/are there taxonomies  that capture, label, define and classify different
> kinds of assumptions that exist in models/modeling?  Ideally with
> discussion of how each (type?) are typically used and more importantly,
> suggested standards for how to report there presence in any/all models.
>
> This is not just a POPMOD issue, but is a basic issue for maturing the
> "science" of computational modeling as a "scientific method" of research.
>
> John
>
> Typed with two thumbs on my iPhone.  (757) 318-0671
>
> The master in the art of living makes little distinction between his work
> and his play, ,,, He simply pursues his vision of excellence at whatever
> he does, leaving others to decide whether he is working or playing."
> James A. Michener (1907 – 1997)
>
> On Oct 2, 2015, at 5:05, popmodwkgrpimag-news-request at simtk.org wrote:
>
> Send PopModWkGrpIMAG-news mailing list submissions to
>    popmodwkgrpimag-news at simtk.org
>
> To subscribe or unsubscribe via the World Wide Web, visit
>    https://simtk.org/mailman/listinfo/popmodwkgrpimag-news
> or, via email, send a message with subject or body 'help' to
>    popmodwkgrpimag-news-request at simtk.org
>
> You can reach the person managing the list at
>    popmodwkgrpimag-news-owner at simtk.org
>
> When replying, please edit your Subject line so it is more specific
> than "Re: Contents of PopModWkGrpIMAG-news digest..."
>
>
> Today's Topics:
>
>   1. Re: Population modeling (Jacob Barhak)
>
>
> ----------------------------------------------------------------------
>
> Message: 1
> Date: Thu, 1 Oct 2015 16:41:20 -0500
> From: Jacob Barhak <jacob.barhak at gmail.com>
> To: Roger Jelliffe <jelliffe at usc.edu>
> Cc: "popmodwkgrpimag-news at simtk.org" <popmodwkgrpimag-news at simtk.org>,
>    "Michael N. Neely" <mneely at usc.edu>
> Subject: Re: [Population Modeling] Population modeling
> Message-ID:
>    <CAM_y+3Q8VjqSYTGXAv7KngW40BnLD0admErTQPBdXyBWzJfrQg at mail.gmail.com>
> Content-Type: text/plain; charset="utf-8"
>
> Thanks Roger,
>
> So indeed, how many modelers in this group use a similar estimation
> approach to Roger?
>
> Please do join this discussion regarding estimation techniques. If enough
> of us do so, we can write a short summary like we did last year - only
> this time focus on estimation techniques.
>
> I hope enough of you will be interested to expose your estimation work for
> population models this way.
>
>                Jacob
>
>
> On Thu, Oct 1, 2015 at 4:33 PM, Roger Jelliffe <jelliffe at usc.edu> wrote:
>
> Dear Jacob:
>
>
>         Thanks for your reply. Here is my two cents again.
>
>
> How many in the group use nonparametric methods for population analysis?
>
> These methods do not make any assumptions as to the presumed shape of the
>
> probability distribution of the model parameters ? they are assumption
>
> free, and there are no constraints at all on what that distribution can be.
>
> The Pmetrics software by the USC Laboratory of Applied Pharmacokinetics and
>
> Bioinformatics makes this software available free over www.lapk.org.
>
> Actually, any system that can be described by ordinary differential
>
> equations can be analyzed with this software, not just
>
> pharmacokinetic/dynamic systems.  In addition, they permit maximally
>
> precise control of the system (dosage regimens of drugs, for example) using
>
> multiple model (MM) dosage design. At the web site, you can also access
>
> publications and technical reports dealing with the subject. For openers,
>
> one might look at these references:
>
>
> 1.    Bustad A, Terziivanov D, Leary R, Port R, Schumitzky A, and
>
> Jelliffe R: Parametric and Nonparametric Population Methods: Their
>
> Comparative Performance in Analysing a Clinical Data Set and Two Monte
>
> Carlo Simulation Studies. Clin. Pharmacokinet., 45: 365-383, 2006.
>
>
> 2.   Neely M, van Guilder M, Yamada W, Schumitzky A, and Jelliffe R:
>
> Accurate Detection of Outliers and Subpopulations with Pmetrics, a
>
> Nonparametric and Parametric Pharmacometric Modeling and Simulation Package
>
> for R. Therap. Drug Monit. 34: 467-476, 2012.
>
>
> 3.   Jelliffe R, Schumitzky A, Bayard D, Milman M, Van Guilder M, Wang X,
>
> Jiang F, Barbaut X, and Maire P: Model-Based, Goal-Oriented, Individualized
>
> Drug Therapy: Linkage of Population Modeling, New "Multiple Model" Dosage
>
> Design, Bayesian Feedback, and Individualized Target Goals. Clin.
>
> Pharmacokinet. 34: 57-77, 1998.
>
>
> Likelihoods are greater with NP models because the distributions are not
>
> constrained by any assumptions about its shape, such as normal, lognormal,
>
> bimodal, etc. The distributions simply are what they are. There is also a
>
> good Monte Carlo simulation routine that can do rigorous MC simulation even
>
> when the basic discrete structure of a nonparametric model with its
>
> multiple discrete support points is preserved.
>
>
> Best regards to all,
>
>
> Roger Jelliffe
>
>
>
>
>
>
>
>
> Roger W. Jelliffe, M.D., F.C.P., F.A.A.C.S.
>
>
> Professor of Medicine Emeritus,
>
>
> USC School of Medicine
>
>
> Founder and Director Emeritus
>
>
> Laboratory of Applied Pharmacokinetics and Bioinformatics
>
>
> Consultant in Infectious Diseases,
>
>
> Children?s Hospital of Los Angeles
>
>
> 4640 Hollywood Blvd, MS #30
>
>
> Los Angeles CA 90027
>
>
> Office - 323-361-5046
>
>
> Fax - 323-361-5045
>
>
> Cell - 626-484-5313
>
>
> jelliffe at usc.edu
>
>
> www.lapk.org
>
>
>
>
> Quantitative approaches
>
>
> to optimally precise individualized drug therapy
>
>
> are more caring,
>
>
> and useful,
>
>
> scientifically, medically and socially,
>
>
> than all the memorized words and facts
>
>
> of categorized and classified experience
>
>
> can ever be!
>
>
>
>
> _______________________________________________
>
> PopModWkGrpIMAG-news mailing list
>
> PopModWkGrpIMAG-news at simtk.org
>
> https://simtk.org/mailman/listinfo/popmodwkgrpimag-news
>
>
>
> -------------- next part --------------
> An HTML attachment was scrubbed...
> URL: <
> https://simtk.org/pipermail/popmodwkgrpimag-news/attachments/20151001/2f30dcb1/attachment-0001.html
> >
>
> ------------------------------
>
> _______________________________________________
> PopModWkGrpIMAG-news mailing list
> PopModWkGrpIMAG-news at simtk.org
> https://simtk.org/mailman/listinfo/popmodwkgrpimag-news
>
>
> End of PopModWkGrpIMAG-news Digest, Vol 12, Issue 5
> ***************************************************
>
>
> _______________________________________________
> PopModWkGrpIMAG-news mailing list
> PopModWkGrpIMAG-news at simtk.org
> https://simtk.org/mailman/listinfo/popmodwkgrpimag-news
>
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://simtk.org/pipermail/popmodwkgrpimag-news/attachments/20151006/560acf71/attachment.html>