<div dir="ltr">Thanks Paul, Thanks Madhav,<br><br> And to all those who cannot make it to the meeting, yet wish to join the<br> discussion, please use this opportunity to add your voices.<br><br> Paul raises questions that create synergy with a previous discussion thread<br> we had when we started this group - especially the first two points Paul<br> uses. Those questions still remain open for discussion. At the same time I<br> would like to remind people of two requests made by our members a few<br> months ago. I am re raising those requests since it seems appropriate to<br> revisit such issues:<br><br> Steve Leff:<br> "... I especially would be interested in hearing from persons who have, or<br> know of persons who have, transition probability data for people with<br> serious mental illness, or persons who have meta-analyzed transition<br> probability data. ..."<br><br> Talitha Feenstra:<br> "... In a new project we want to use this same model as the basis for<br> developing region specific models that add up to a national model. And<br> think about smart ways to do so, using region specific data when available,<br> but national data when needed. I would be interested in discussing some of<br> the issues met in that project with the group members. ..."<br><br> Both Steve and Talitha are touching upon issues that are of recent interest:<br> Data availability and scale levels. The former point was raised recently by<br> Madhav and led to some communications in our group. And the latter may be<br> especially interesting in for the Multi Scale modeling consortium.<br><br> Paul, I would appreciate it if you add those points to our discussions.<br><br><br> Jacob</div><div class="gmail_extra"><br><div class="gmail_quote">On Mon, Aug 24, 2015 at 5:38 PM, Madhav Marathe <span dir="ltr"><<a href="mailto:mmarathe@vbi.vt.edu" target="_blank">mmarathe@vbi.vt.edu</a>></span> wrote:<br><blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex">Dear Paul<br>
<br>
Thanks a lot. Appreciate it. These are very good questions and capture our theme very well. As Paul has said, we welcome<br>
additional comments and suggestions.<br>
<span class="HOEnZb"><font color="#888888"><br>
Madhav<br>
</font></span><div class="HOEnZb"><div class="h5"><br>
----- Original Message -----<br>
From: "Paul Marjoram" <<a href="mailto:pmarjora@usc.edu">pmarjora@usc.edu</a>><br>
To: "Jacob Barhak" <<a href="mailto:jacob.barhak@gmail.com">jacob.barhak@gmail.com</a>><br>
Cc: <a href="mailto:popmodwkgrpimag-news@simtk.org">popmodwkgrpimag-news@simtk.org</a><br>
Sent: Monday, August 24, 2015 6:27:22 PM<br>
Subject: Re: [Population Modeling] MSM/IMAG meeting on Sep 8-10<br>
<br>
Thanks Jacob. Let me tag onto this message to say that there will be a Breakout Session for the Population Group on Day 1, which I believe Madhav and I will be leading.<br>
We’d like to solicit opinions regarding what the agenda of that meeting should be. One thought is to focus the discussion around a number of topics related to population modeling. Here (below) are a collection of topics that we might feature in such a discussion. Please do chime in if you would like to add or subtract from that least, or if you have other ideas that you would like to propose for that Breakout meeting.<br>
<br>
Suggested Discussion Themes For Breakout Session for Population Working Group:<br>
1. Is heterogeneity modeled? If so, how is it represented? Are the individuals heterogeneous or does the heterogeneity appear through their exposures/environment? Is there a fundamental difference between heterogeneity in the individuals or their environment? When is it important to reflect the heterogeneity that is present in virtually any population, and when can that heterogeneity be safely ignored?<br>
2. Is the model at the individual (agent) or population level? How do we choose which level is the most appropriate at which to build the model. How (or are) these two levels tied together? Models at individual level seem to naturally imply population level behavior, but not vice-versa.<br>
3. How is the model performance assessed? Is model fit addressed formally? I.e. how do we decide whether a model is actually useful? What principles are used and how are they chosen? Is this addressed in an ad-hoc fashion or are formal statistical methods used? How is the simulation data fit to the observed data?<br>
4. This one is an actual theme. By far the most common application theme is human health, and typically epidemiology. This makes sense, since humans are the population about which we care most. We might then ask when lessons learned from human populations can be usefully applied to other populations? Also, what features makes humans a particularly challenging population to model?<br>
<br>
<br>
Thanks, Paul<br>
<br>
<br>
<br>
<br>
<br>
On Aug 24, 2015, at 3:21 PM, Jacob Barhak < <a href="mailto:jacob.barhak@gmail.com">jacob.barhak@gmail.com</a> > wrote:<br>
<br>
Hello All,<br>
<br>
This is a reminder that the MSM/IMAG meeting will take place at the NIH on Sep 8-10. This is the meeting where the population working group met which led to the start of this mailing list. And there will be another meeting this year.<br>
<br>
An agenda was drafted on the wiki and can be found at:<br>
<a href="http://www.imagwiki.nibib.nih.gov/imag-events/2015-msm-consortium-meeting" target="_blank" rel="noreferrer">http://www.imagwiki.nibib.nih.gov/imag-events/2015-msm-consortium-meeting</a><br>
<br>
Last time I checked, there were still tickets and you can still register though this link:<br>
<a href="https://www.eventbrite.com/e/2015-imag-multiscale-modeling-consortium-tickets-17544718726" target="_blank" rel="noreferrer">https://www.eventbrite.com/e/2015-imag-multiscale-modeling-consortium-tickets-17544718726</a><br>
<br>
If you are interested, you may wish to act fast while tickets remain.<br>
<br>
If you know you are going, please let us know so members of this group can find each other easily - I know Madhav, Paul Marjoram, And Tony Hunt will be there. Hopefully more of you that can travel easily will appear.<br>
<br>
This event is highly recommended for medical computational modelers especially due to its collaborative nature.<br>
<br>
Jacob<br>
<br>
_______________________________________________<br>
PopModWkGrpIMAG-news mailing list<br>
<a href="mailto:PopModWkGrpIMAG-news@simtk.org">PopModWkGrpIMAG-news@simtk.org</a><br>
<a href="https://simtk.org/mailman/listinfo/popmodwkgrpimag-news" target="_blank" rel="noreferrer">https://simtk.org/mailman/listinfo/popmodwkgrpimag-news</a><br>
<br>
<br>
<br>
=====================<br>
=====================<br>
Paul Marjoram<br>
Research Assoc. Prof.<br>
Division of Biostatistics<br>
U.S.C., Dept. of Prev. Med.<br>
2001 N. Soto St.,<br>
SSB 202V, MC-9234<br>
Los Angeles<br>
CA 90089-9234<br>
<a href="mailto:pmarjora@usc.edu">pmarjora@usc.edu</a><br>
323-442-0111<br>
<br>
<br>
<br>
<br>
<br>
_______________________________________________<br>
PopModWkGrpIMAG-news mailing list<br>
<a href="mailto:PopModWkGrpIMAG-news@simtk.org">PopModWkGrpIMAG-news@simtk.org</a><br>
<a href="https://simtk.org/mailman/listinfo/popmodwkgrpimag-news" target="_blank" rel="noreferrer">https://simtk.org/mailman/listinfo/popmodwkgrpimag-news</a><br>
</div></div></blockquote></div><br></div>