Guidelines for a fresh modeller

5 messages 4 people Latest: Nov 25, 2011

Guidelines for a fresh modeller

From: Elisabet Størset Date: November 24, 2011 technical
Dear NONMEMusers I am working with my very first pop-PK model with NONMEM and have no experience. I have read a lot of theory but lack experience of actually using nonmem. I see the OFV and estimates, but have trouble knowing how to interpret the results and the graphs, and how to improve the structural model before incorporate covariates. When I have nothing to compare with, it is difficult to know if I have found the right structural model and whether my results are actually good or bad. I also wonder what kind of plots in the plot jungle which you experience modelers find most useful to look at when trying to improve a model, and how you know which starting estimates to choose. Maybe any of you remember how it was to be a fresh modeller, and remember what you did to get your first experience. I could use a good, pedagogical, stepwise guide on modelling, with some illustrative examples and excercises to practice with before starting on my own dataset. Do any of you know if this is available? Thanks in advance :)

RE: Guidelines for a fresh modeller

From: Francois Gaudreault Date: November 24, 2011 technical
Hi Elisabet When I first started, I found very usefull the ACCP web site where you have lots of examples and tutorials. Here is the link: http://accp1.org/pharmacometrics/cssolutionnm.htm Please also consider boomer.org and finally Fisher/Shafer NONMEM Workshop, all freely available on the internet Good luck -- François Gaudreault, Ph.D. Candidate Pharmacométrie / Pharmacometrics Charger de Cours / Lecturer Président / President Conseil étudiant cycles supérieurs / Graduate Students Association Faculté de pharmacie / Faculty of Pharmacy Université de Montréal Date: Thu, 24 Nov 2011 16:54:56 +0100 Subject: [NMusers] Guidelines for a fresh modeller From: [email protected] To: [email protected] Dear NONMEMusers I am working with my very first pop-PK model with NONMEM and have no experience. I have read a lot of theory but lack experience of actually using nonmem. I see the OFV and estimates, but have trouble knowing how to interpret the results and the graphs, and how to improve the structural model before incorporate covariates. When I have nothing to compare with, it is difficult to know if I have found the right structural model and whether my results are actually good or bad. I also wonder what kind of plots in the plot jungle which you experience modelers find most useful to look at when trying to improve a model, and how you know which starting estimates to choose. Maybe any of you remember how it was to be a fresh modeller, and remember what you did to get your first experience. I could use a good, pedagogical, stepwise guide on modelling, with some illustrative examples and excercises to practice with before starting on my own dataset. Do any of you know if this is available? Thanks in advance :)

Guidelines for a fresh modeller

From: Carlos Hoyo Date: November 24, 2011 technical
Dear Elisabeth, You can compare the fits for one compartment versus two compartments and you can introduce little by little minor changes to the model to make the comparisons. Carlos Hoyo Farma, Cinvestav Mexico City

Re: Guidelines for a fresh modeller

From: Brian Scott Decker Date: November 24, 2011 technical
I would also be interested in this. Thank-you. Brian Decker MD, PharmD Elisabet Størset <[email protected]> wrote: Dear NONMEMusers I am working with my very first pop-PK model with NONMEM and have no experience. I have read a lot of theory but lack experience of actually using nonmem. I see the OFV and estimates, but have trouble knowing how to interpret the results and the graphs, and how to improve the structural model before incorporate covariates. When I have nothing to compare with, it is difficult to know if I have found the right structural model and whether my results are actually good or bad. I also wonder what kind of plots in the plot jungle which you experience modelers find most useful to look at when trying to improve a model, and how you know which starting estimates to choose. Maybe any of you remember how it was to be a fresh modeller, and remember what you did to get your first experience. I could use a good, pedagogical, stepwise guide on modelling, with some illustrative examples and excercises to practice with before starting on my own dataset. Do any of you know if this is available? Thanks in advance :)

Re: Guidelines for a fresh modeller

From: Elisabet Størset Date: November 25, 2011 technical
Thanks to everyone for your very kind answers! It will help me getting started. My experience with the modelling world so far is that interest in modelling correlates with being friendly and helpful. :)
Quoted reply history
2011/11/24 Carlos Hoyo <[email protected]> > Dear Elisabeth, > > You can compare the fits for one compartment versus two compartments and > you can introduce little by little minor changes to the model to make the > comparisons. > > Carlos Hoyo > Farma, Cinvestav > Mexico City