[NMusers] TR: Post doctoral postion in Sanofi R&D Paris

[Zsolt.Bocskei]

Dear Colleagues,


A post-doctoral fellowship is available for a period of 12 months in the 
department of Structure, Design & Informatics at Sanofi R&D, Vitry-sur-Seine, 
France for a successful applicant with a recent PhD or up to 3 years 
postdoctoral experience in systems biology.

The main missions of the postdoctoral fellowship are as follows:

1./Develop and internalize mechanistic mathematical models to signaling 
networks of interest for the oncology field (1,2)

2./Contribute to the precise description and understanding of the selected drug 
target networks and their interactions with small molecular or biologic drugs

3./Participation in mechanistic PK/PD simulations to drug discovery projects.

We would like to develop, validate and use models around different cell lines 
of interest for Oncology applications. We also intend to investigate the effect 
of the inhibition of different feedback and feed forward mechanisms as well as 
the effect of combination treatments.

Experience in the computational systems biology field, understanding of basic 
pharmacology concepts and good skills in mathematics as well as knowledge of 
modeling tools like Matlab/Simbiology, Copasi or Celldesigner etc are 
prerequisites.

Good communication skills, ability to work in a collaborative fashion and 
proficiency in English are a must.
Sanofi is a global pharmaceutical company with more than 110 000 staff 
worldwide. Our research site in Vitry-sur-Seine (15 km from the center of
Paris) employs more than 1000 people and offers a stimulating, 
interdisciplinary research environment.

(1) Klinger et al Mol. Sys Biol. 9:673 (2013)
(2) Chen et al Mol. Sys. Biol. 5:239 (2009)

In case of interest please apply via our company web site:

French

https://fr-fr.sanofi-aventis-job.com/ERECRUITMENT/(S(tubdwu45nvsjuu55emptwjnx))/pages/Jobs/JobDetail.aspx?lang=fr-fr&Job_Id=ubG5IaOVsHA%24

English

https://fr-en.sanofi-aventis-job.com/ERECRUITMENT/(S(rxtqe155w5hegs45na12w155))/pages/Jobs/JobDetail.aspx?lang=en-fr&Job_Id=ubG5IaOVsHA%24

For more information please contact 
zsolt.bocs...@sanofi.com.


Best regards,

Zsolt Böcskei
Structure Design and Informatics
Lead Generation and Candidate Realization
TEL.: +33 (0) 1.53.77.45.08  -  CELL.: +33 (0) 6.32.32.67.10
13, Quai Jules Guesde - 94403 Vitry-sur-Seine - France

[cid:image001.jpg@01CEB526.7F33D0D0]

Please consider the environment before printing this email!

<>

[NMusers] REGISTRATION to Annual Indiana CTSI Symposium is OPEN

[Chitnis, Shripad]

Registration to the 3rd Annual Indiana CTSI Symposium is now OPEN !
……..
The Third Annual Indiana Clinical and Translational Sciences Institute (CTSI) 
Symposium on Disease and Therapeutic Response Modeling
November 5 and 6, 2013
We are pleased to announce that the third annual Indiana Clinical and 
Translational Sciences Institute (CTSI) Symposium on Disease and Therapeutic 
Response Modeling will take place on November 5-6, 2013 in Indianapolis, 
Indiana. This two-day program will include speakers from academia, industry and 
regulatory agencies, as well as poster presentations. The confirmed speakers 
include – Dr. Vikram Sinha (FDA), Dr. Jeffrey Barrett (Children's Hospital of 
Philadelphia), Dr. Marc Gastonguay (Metrum Institute), Dr. John Urquhart 
(AARDEX Group), Dr. Mark Sale (Next Level Solutions, LLC), Dr. Sean Mooney 
(Buck Institute for Age Research), Dr. Michael Heathman (Eli Lilly and 
Company), and Dr. Bernard Vrijens (MeadWestvaco Healthcare).

REGISTRATION

Link to registration:  
https://www.indianactsi.org/jevents/eventdetail/7/-/third-annual-indiana-clinical-and-translational-sciences-institute-symposium-on-disease-and-therapeutic-response-modeling

Academic, Government and Eli Lilly Company Free
Industry $200

CONFERENCE LOCATION:

Campus Center
420 University Blvd. CE 278
Indianapolis, IN, 46202
Phone:(317) 274-

For more information or questions, please contact:
Shripad Chitnis (chitn...@iupui.edu) or Nieves Velez 
de Mendizabal (nvele...@iu.edu)

We look forward to seeing you in November!

Best Regards,
Shripad Chitnis
Fellow, Division of Clinical Pharmacology,
Indiana University School of Medicine
Indiana CTSI Disease and Therapeutic Response Modeling Program

[NMusers] Parameter Estimation in IF Conditioning Statement

[benjamin.weber]

Dear NONMEM users,

My goal was to add a drug effect to a disease progression model in a way that 
the drug effect is different when a certain threshold concentration is exceeded 
(similarly to modeling a 'hockey-stick' when performing covariate analysis). 
First, I did it in the following way (partial simplified code)

$PRED

THRESH = THETA(1) ;threshold concentration

INT = THETA(2)+ ETA(1)

IF(WCMIN.LE.THRESH) SLOPE = THETA(3)+THETA(4)*CONC+ETA(2)
IF(WCMIN.GT.THRESH) SLOPE = THETA(3)+THETA(5)*CONC+ETA(2)

IPRED=INT+SLOPE*TIME

and NONMEM did not manage to estimate THRESH=THETA(1) (i.e., the initial 
estimate did not change during minimization and the gradient was zero 
throughout). I tried this for several different initial estimates.

I then centered the observed CONC on THRESH and run the following model because 
colleagues mentioned that they have successfully run a 'hockey-stick' 
estimating the threshold parameter  (partial simplified code)

IF(WCMIN.LE.THRESH) SLOPE = THETA(3)+THETA(4)*(CONC-THRESH)+ETA(2)
IF(WCMIN.GT.THRESH) SLOPE = THETA(3)+THETA(5)*(CONC- THRESH)+ETA(2)

This time, NONMEM provided very reasonable estimates for all model parameters 
and the covariance step was successful.

I wonder now if the centering of the variable (and hence something particular 
to the data set) caused the difference in estimability of the threshold 
parameter or whether NONMEN cannot estimate parameters that only occur in the 
conditioning part of the IF statement (Note that after centering THRESH also 
appears in another part of the code).

Could somebody please provide some insight on this? I have searched in the 
NONMEM user group but could not find anything.


Mit freundlichen Grüßen / Kind regards,
Dr. Benjamin Weber

Boehringer Ingelheim Pharma GmbH & Co. KG
Translational Medicine
Tel.: +49 (7351) 54-143520
Fax: +49 (7351) 83-143520
mailto:benjamin.we...@boehringer-ingelheim.com

Boehringer Ingelheim Pharma GmbH & Co. KG, Sitz: Ingelheim am Rhein; 
Registergericht Mainz: HR A 22206; Komplementär Boehringer Ingelheim 
Deutschland GmbH; Geschäftsführung: Dr. Engelbert Günster (Vorsitzender),  
Ursula Fuggis-Hahn, Ralf Gorniak,  Michael Klein, Dr. Martin Wanning; 
Vorsitzender des Aufsichtsrates: Dr. Joachim Hasenmaier; Sitz: Ingelheim am 
Rhein; Registergericht Mainz: HR B 23260

Diese E-Mail ist vertraulich zu behandeln. Sie kann besonderem rechtlichem 
Schutz unterliegen. Wenn Sie nicht der richtige Adressat sind, senden Sie bitte 
diese E-Mail an den Absender zurück, löschen die eingegangene E-Mail und geben 
den Inhalt der E-Mail nicht weiter. Jegliche unbefugte Bearbeitung, Nutzung, 
Vervielfältigung oder Verbreitung ist verboten. / This e-mail is confidential 
and may also be legally privileged. If you are not the intended recipient 
please reply to sender, delete the e-mail and do not disclose its contents to 
any person. Any unauthorized review, use, disclosure, copying or distribution 
is strictly prohibited.

Re: [NMusers] Parameter Estimation in IF Conditioning Statement

[Leonid Gibiansky]

Hi Benjamin
I think the problem was that when nonmem tested small variations of 
THETA(1) in the first version, nothing changed in the OF as these 
variation have not resulted in the change in the IF conditions (e.g., if 
your data contained WCMIN values 1 and 2 and nothing in between, and 
THETA(1) was 1.5, changes of THETA(1) to 1.6 or 1.4 were not resulting 
in the changes in OF, gradient was zero, and the gradient method was 
unable to move the model. When you put theta(1) in the centering, this 
resulted in changes of OF. This is not an unusual behavior. Therefore, 
it is better to use continuous functions rather than switches. For 
example, you can code it as


INT = THETA(4)+(THETA(5)-THETA(4))/(1+(THETA(1)/WCMIN)**GAM)

SLOPE = THETA(3)+INT*CONC+ETA(2)

When GAM is infinity, this is equivalent to your model. You may use 
large GAM (or even estimated GAM)


Leonid



--
Leonid Gibiansky, Ph.D.
President, QuantPharm LLC
web:www.quantpharm.com
e-mail: LGibiansky at quantpharm.com
tel:(301) 767 5566



On 9/19/2013 2:31 PM, benjamin.we...@boehringer-ingelheim.com wrote:

Dear NONMEM users,

My goal was to add a drug effect to a disease progression model in a way
that the drug effect is different when a certain threshold concentration
is exceeded (similarly to modeling a ‘hockey-stick’ when performing
covariate analysis). First, I did it in the following way (partial
simplified code)

$PRED

THRESH = THETA(1) ;threshold concentration

INT = THETA(2)+ ETA(1)

IF(WCMIN.LE.THRESH) SLOPE = THETA(3)+THETA(4)*CONC+ETA(2)

IF(WCMIN.GT.THRESH) SLOPE = THETA(3)+THETA(5)*CONC+ETA(2)

IPRED=INT+SLOPE*TIME

and NONMEM did not manage to estimate THRESH=THETA(1) (i.e., the initial
estimate did not change during minimization and the gradient was zero
throughout). I tried this for several different initial estimates.

I then centered the observed CONC on THRESH and run the following model
because colleagues mentioned that they have successfully run a
‘hockey-stick’ estimating the threshold parameter  (partial simplified code)

IF(WCMIN.LE.THRESH) SLOPE = THETA(3)+THETA(4)*(CONC-THRESH)+ETA(2)

IF(WCMIN.GT.THRESH) SLOPE = THETA(3)+THETA(5)*(CONC- THRESH)+ETA(2)

This time, NONMEM provided very reasonable estimates for all model
parameters and the covariance step was successful.

I wonder now if the centering of the variable (and hence something
particular to the data set) caused the difference in estimability of the
threshold parameter or whether NONMEN cannot estimate parameters that
only occur in the conditioning part of the IF statement (Note that after
centering THRESH also appears in another part of the code).

Could somebody please provide some insight on this? I have searched in
the NONMEM user group but could not find anything.

Mit freundlichen Grüßen / Kind regards,
Dr. Benjamin Weber

Boehringer Ingelheim Pharma GmbH & Co. KG
Translational Medicine
Tel.: +49 (7351) 54-143520
Fax: +49 (7351) 83-143520
mailto:benjamin.we...@boehringer-ingelheim.com

Boehringer Ingelheim Pharma GmbH & Co. KG, Sitz: Ingelheim am Rhein;
Registergericht Mainz: HR A 22206; Komplementär Boehringer Ingelheim
Deutschland GmbH; Geschäftsführung: Dr. Engelbert Günster
(Vorsitzender),  Ursula Fuggis-Hahn, Ralf Gorniak,  Michael Klein, Dr.
Martin Wanning; Vorsitzender des Aufsichtsrates: Dr. Joachim Hasenmaier;
Sitz: Ingelheim am Rhein; Registergericht Mainz: HR B 23260

Diese E-Mail ist vertraulich zu behandeln. Sie kann besonderem
rechtlichem Schutz unterliegen. Wenn Sie nicht der richtige Adressat
sind, senden Sie bitte diese E-Mail an den Absender zurück, löschen die
eingegangene E-Mail und geben den Inhalt der E-Mail nicht weiter.
Jegliche unbefugte Bearbeitung, Nutzung, Vervielfältigung oder
Verbreitung ist verboten. / This e-mail is confidential and may also be
legally privileged. If you are not the intended recipient please reply
to sender, delete the e-mail and do not disclose its contents to any
person. Any unauthorized review, use, disclosure, copying or
distribution is strictly prohibited.

AW: Re: [NMusers] Parameter Estimation in IF Conditioning Statement

[benjamin.weber]

Hi Leonid. Thanks for your quick response. WCMIN is continuous. In fact, it is 
actually identical to CONC (forgot to change it while simplifying). Any 
thoughts on this?Ben

- Originalnachricht -
Von: Leonid Gibiansky [mailto:lgibian...@quantpharm.com]
Gesendet: Thursday, September 19, 2013 09:32 PM
An: Weber,Dr.,Benjamin (TransMed) BIP-DE-B
Cc: nmusers@globomaxnm.com 
Betreff: Re: [NMusers] Parameter Estimation in IF Conditioning Statement

Hi Benjamin
I think the problem was that when nonmem tested small variations of 
THETA(1) in the first version, nothing changed in the OF as these 
variation have not resulted in the change in the IF conditions (e.g., if 
your data contained WCMIN values 1 and 2 and nothing in between, and 
THETA(1) was 1.5, changes of THETA(1) to 1.6 or 1.4 were not resulting 
in the changes in OF, gradient was zero, and the gradient method was 
unable to move the model. When you put theta(1) in the centering, this 
resulted in changes of OF. This is not an unusual behavior. Therefore, 
it is better to use continuous functions rather than switches. For 
example, you can code it as

INT = THETA(4)+(THETA(5)-THETA(4))/(1+(THETA(1)/WCMIN)**GAM)

SLOPE = THETA(3)+INT*CONC+ETA(2)

When GAM is infinity, this is equivalent to your model. You may use 
large GAM (or even estimated GAM)

Leonid



--
Leonid Gibiansky, Ph.D.
President, QuantPharm LLC
web:www.quantpharm.com
e-mail: LGibiansky at quantpharm.com
tel:(301) 767 5566



On 9/19/2013 2:31 PM, benjamin.we...@boehringer-ingelheim.com wrote:
> Dear NONMEM users,
>
> My goal was to add a drug effect to a disease progression model in a way
> that the drug effect is different when a certain threshold concentration
> is exceeded (similarly to modeling a ‘hockey-stick’ when performing
> covariate analysis). First, I did it in the following way (partial
> simplified code)
>
> $PRED
>
> THRESH = THETA(1) ;threshold concentration
>
> INT = THETA(2)+ ETA(1)
>
> IF(WCMIN.LE.THRESH) SLOPE = THETA(3)+THETA(4)*CONC+ETA(2)
>
> IF(WCMIN.GT.THRESH) SLOPE = THETA(3)+THETA(5)*CONC+ETA(2)
>
> IPRED=INT+SLOPE*TIME
>
> and NONMEM did not manage to estimate THRESH=THETA(1) (i.e., the initial
> estimate did not change during minimization and the gradient was zero
> throughout). I tried this for several different initial estimates.
>
> I then centered the observed CONC on THRESH and run the following model
> because colleagues mentioned that they have successfully run a
> ‘hockey-stick’ estimating the threshold parameter  (partial simplified code)
>
> IF(WCMIN.LE.THRESH) SLOPE = THETA(3)+THETA(4)*(CONC-THRESH)+ETA(2)
>
> IF(WCMIN.GT.THRESH) SLOPE = THETA(3)+THETA(5)*(CONC- THRESH)+ETA(2)
>
> This time, NONMEM provided very reasonable estimates for all model
> parameters and the covariance step was successful.
>
> I wonder now if the centering of the variable (and hence something
> particular to the data set) caused the difference in estimability of the
> threshold parameter or whether NONMEN cannot estimate parameters that
> only occur in the conditioning part of the IF statement (Note that after
> centering THRESH also appears in another part of the code).
>
> Could somebody please provide some insight on this? I have searched in
> the NONMEM user group but could not find anything.
>
> Mit freundlichen Grüßen / Kind regards,
> Dr. Benjamin Weber
>
> Boehringer Ingelheim Pharma GmbH & Co. KG
> Translational Medicine
> Tel.: +49 (7351) 54-143520
> Fax: +49 (7351) 83-143520
> mailto:benjamin.we...@boehringer-ingelheim.com
>
> Boehringer Ingelheim Pharma GmbH & Co. KG, Sitz: Ingelheim am Rhein;
> Registergericht Mainz: HR A 22206; Komplementär Boehringer Ingelheim
> Deutschland GmbH; Geschäftsführung: Dr. Engelbert Günster
> (Vorsitzender),  Ursula Fuggis-Hahn, Ralf Gorniak,  Michael Klein, Dr.
> Martin Wanning; Vorsitzender des Aufsichtsrates: Dr. Joachim Hasenmaier;
> Sitz: Ingelheim am Rhein; Registergericht Mainz: HR B 23260
>
> Diese E-Mail ist vertraulich zu behandeln. Sie kann besonderem
> rechtlichem Schutz unterliegen. Wenn Sie nicht der richtige Adressat
> sind, senden Sie bitte diese E-Mail an den Absender zurück, löschen die
> eingegangene E-Mail und geben den Inhalt der E-Mail nicht weiter.
> Jegliche unbefugte Bearbeitung, Nutzung, Vervielfältigung oder
> Verbreitung ist verboten. / This e-mail is confidential and may also be
> legally privileged. If you are not the intended recipient please reply
> to sender, delete the e-mail and do not disclose its contents to any
> person. Any unauthorized review, use, disclosure, copying or
> distribution is strictly prohibited.
>

Re: AW: Re: [NMusers] Parameter Estimation in IF Conditioning Statement

[Leonid Gibiansky]

Benjamin.
Even as WCMIN is continuous by nature, it is not continuous in the 
dataset (at observation points). You may look at your WCMIN values in 
the data file (I assume this values are in the data file?) and see where 
the initial conditions for THETA(1) placed in that range. Step for 
gradient computations can be quite small, so it will not see continuous 
nature of the variable, and will treat it as discrete.


One the same model but separate topic, I am not sure that you can code 
the model as you coded it if you indeed assume that WCMIN=CONC.

It is better to write as (set aside problems with the gradient):

IF(WCMIN.LE.THRESH) SLOPE = THETA(3)+THETA(4)*CONC+ETA(2)

IF(WCMIN.GT.THRESH) SLOPE = THETA(3)*WCMIN+THETA(5)*(CONC-WCMIN)+ETA(2)

Otherwise, you SLOP function will be discontinuous at WCMIN as a 
function of CONC.


Note also that your two models are not exactly equivalent. If you use
IF(WCMIN.LE.THRESH) SLOPE = THETA(3)+THETA(4)*CONC+ETA(2)
IF(WCMIN.GT.THRESH) SLOPE = THETA(3)+THETA(5)*CONC+ETA(2)

your values at WCMIN=THRESH are equal to
SLOPE = THETA(3)+THETA(4)*THRESH+ETA(2)
SLOPE = THETA(3)+THETA(5)*THRESH+ETA(2)
immediately before and after WCMIN=THRESH (different)

while for the second model they are equal to

SLOPE = THETA(3)+ETA(2)
SLOPE = THETA(3)+ETA(2)

and equal to each other

Leonid


--
Leonid Gibiansky, Ph.D.
President, QuantPharm LLC
web:www.quantpharm.com
e-mail: LGibiansky at quantpharm.com
tel:(301) 767 5566



On 9/19/2013 4:29 PM, benjamin.we...@boehringer-ingelheim.com wrote:

Hi Leonid. Thanks for your quick response. WCMIN is continuous. In fact, it is 
actually identical to CONC (forgot to change it while simplifying). Any 
thoughts on this?Ben

- Originalnachricht -
Von: Leonid Gibiansky [mailto:lgibian...@quantpharm.com]
Gesendet: Thursday, September 19, 2013 09:32 PM
An: Weber,Dr.,Benjamin (TransMed) BIP-DE-B
Cc: nmusers@globomaxnm.com 
Betreff: Re: [NMusers] Parameter Estimation in IF Conditioning Statement

Hi Benjamin
I think the problem was that when nonmem tested small variations of
THETA(1) in the first version, nothing changed in the OF as these
variation have not resulted in the change in the IF conditions (e.g., if
your data contained WCMIN values 1 and 2 and nothing in between, and
THETA(1) was 1.5, changes of THETA(1) to 1.6 or 1.4 were not resulting
in the changes in OF, gradient was zero, and the gradient method was
unable to move the model. When you put theta(1) in the centering, this
resulted in changes of OF. This is not an unusual behavior. Therefore,
it is better to use continuous functions rather than switches. For
example, you can code it as

INT = THETA(4)+(THETA(5)-THETA(4))/(1+(THETA(1)/WCMIN)**GAM)

SLOPE = THETA(3)+INT*CONC+ETA(2)

When GAM is infinity, this is equivalent to your model. You may use
large GAM (or even estimated GAM)

Leonid



--
Leonid Gibiansky, Ph.D.
President, QuantPharm LLC
web:www.quantpharm.com
e-mail: LGibiansky at quantpharm.com
tel:(301) 767 5566



On 9/19/2013 2:31 PM, benjamin.we...@boehringer-ingelheim.com wrote:

Dear NONMEM users,

My goal was to add a drug effect to a disease progression model in a way
that the drug effect is different when a certain threshold concentration
is exceeded (similarly to modeling a ‘hockey-stick’ when performing
covariate analysis). First, I did it in the following way (partial
simplified code)

$PRED

THRESH = THETA(1) ;threshold concentration

INT = THETA(2)+ ETA(1)

IF(WCMIN.LE.THRESH) SLOPE = THETA(3)+THETA(4)*CONC+ETA(2)

IF(WCMIN.GT.THRESH) SLOPE = THETA(3)+THETA(5)*CONC+ETA(2)

IPRED=INT+SLOPE*TIME

and NONMEM did not manage to estimate THRESH=THETA(1) (i.e., the initial
estimate did not change during minimization and the gradient was zero
throughout). I tried this for several different initial estimates.

I then centered the observed CONC on THRESH and run the following model
because colleagues mentioned that they have successfully run a
‘hockey-stick’ estimating the threshold parameter  (partial simplified code)

IF(WCMIN.LE.THRESH) SLOPE = THETA(3)+THETA(4)*(CONC-THRESH)+ETA(2)

IF(WCMIN.GT.THRESH) SLOPE = THETA(3)+THETA(5)*(CONC- THRESH)+ETA(2)

This time, NONMEM provided very reasonable estimates for all model
parameters and the covariance step was successful.

I wonder now if the centering of the variable (and hence something
particular to the data set) caused the difference in estimability of the
threshold parameter or whether NONMEN cannot estimate parameters that
only occur in the conditioning part of the IF statement (Note that after
centering THRESH also appears in another part of the code).

Could somebody please provide some insight on this? I have searched in
the NONMEM user group but could not find anything.

Mit freundlichen Grüßen / Kind regards,
Dr. Benjamin Weber

Boehringer Ingelheim Pharma GmbH & Co. KG
Translational Medicine
Tel.: +49 (7351) 54-143520
Fax

Re: AW: Re: [NMusers] Parameter Estimation in IF Conditioning Statement

[Leonid Gibiansky]

Sorry, there was a typos in the code, here is the version with 
continuous SLOPE function


IF(CONC.LE.THRESH) SLOPE = THETA(3)+THETA(4)*CONC+ETA(2)
IF(CONC.GT.THRESH) SLOPE = 
THETA(3)+THETA(4)*THRESH+THETA(5)*(CONC-THRESH)+ETA(2)


Then, if you rename
NEWTHETA3 = THETA(3)+THETA(4)*THRESH
you will have

IF(CONC.LE.THRESH) SLOPE = NEWTHETA3+THETA(4)*(CONC-THRESH)+ETA(2)
IF(CONC.GT.THRESH) SLOPE = NEWTHETA3+THETA(5)*(CONC-THRESH)+ETA(2)

exactly as your second version (up to the parametrization).

Leonid


--
Leonid Gibiansky, Ph.D.
President, QuantPharm LLC
web:www.quantpharm.com
e-mail: LGibiansky at quantpharm.com
tel:(301) 767 5566



On 9/19/2013 6:42 PM, Leonid Gibiansky wrote:

Benjamin.
Even as WCMIN is continuous by nature, it is not continuous in the
dataset (at observation points). You may look at your WCMIN values in
the data file (I assume this values are in the data file?) and see where
the initial conditions for THETA(1) placed in that range. Step for
gradient computations can be quite small, so it will not see continuous
nature of the variable, and will treat it as discrete.

One the same model but separate topic, I am not sure that you can code
the model as you coded it if you indeed assume that WCMIN=CONC.
It is better to write as (set aside problems with the gradient):

IF(WCMIN.LE.THRESH) SLOPE = THETA(3)+THETA(4)*CONC+ETA(2)

IF(WCMIN.GT.THRESH) SLOPE = THETA(3)*WCMIN+THETA(5)*(CONC-WCMIN)+ETA(2)

Otherwise, you SLOP function will be discontinuous at WCMIN as a
function of CONC.

Note also that your two models are not exactly equivalent. If you use
IF(WCMIN.LE.THRESH) SLOPE = THETA(3)+THETA(4)*CONC+ETA(2)
IF(WCMIN.GT.THRESH) SLOPE = THETA(3)+THETA(5)*CONC+ETA(2)

your values at WCMIN=THRESH are equal to
SLOPE = THETA(3)+THETA(4)*THRESH+ETA(2)
SLOPE = THETA(3)+THETA(5)*THRESH+ETA(2)
immediately before and after WCMIN=THRESH (different)

while for the second model they are equal to

SLOPE = THETA(3)+ETA(2)
SLOPE = THETA(3)+ETA(2)

and equal to each other

Leonid


--
Leonid Gibiansky, Ph.D.
President, QuantPharm LLC
web:www.quantpharm.com
e-mail: LGibiansky at quantpharm.com
tel:(301) 767 5566



On 9/19/2013 4:29 PM, benjamin.we...@boehringer-ingelheim.com wrote:

Hi Leonid. Thanks for your quick response. WCMIN is continuous. In
fact, it is actually identical to CONC (forgot to change it while
simplifying). Any thoughts on this?Ben

- Originalnachricht -
Von: Leonid Gibiansky [mailto:lgibian...@quantpharm.com]
Gesendet: Thursday, September 19, 2013 09:32 PM
An: Weber,Dr.,Benjamin (TransMed) BIP-DE-B
Cc: nmusers@globomaxnm.com 
Betreff: Re: [NMusers] Parameter Estimation in IF Conditioning Statement

Hi Benjamin
I think the problem was that when nonmem tested small variations of
THETA(1) in the first version, nothing changed in the OF as these
variation have not resulted in the change in the IF conditions (e.g., if
your data contained WCMIN values 1 and 2 and nothing in between, and
THETA(1) was 1.5, changes of THETA(1) to 1.6 or 1.4 were not resulting
in the changes in OF, gradient was zero, and the gradient method was
unable to move the model. When you put theta(1) in the centering, this
resulted in changes of OF. This is not an unusual behavior. Therefore,
it is better to use continuous functions rather than switches. For
example, you can code it as

INT = THETA(4)+(THETA(5)-THETA(4))/(1+(THETA(1)/WCMIN)**GAM)

SLOPE = THETA(3)+INT*CONC+ETA(2)

When GAM is infinity, this is equivalent to your model. You may use
large GAM (or even estimated GAM)

Leonid



--
Leonid Gibiansky, Ph.D.
President, QuantPharm LLC
web:www.quantpharm.com
e-mail: LGibiansky at quantpharm.com
tel:(301) 767 5566



On 9/19/2013 2:31 PM, benjamin.we...@boehringer-ingelheim.com wrote:

Dear NONMEM users,

My goal was to add a drug effect to a disease progression model in a way
that the drug effect is different when a certain threshold concentration
is exceeded (similarly to modeling a ‘hockey-stick’ when performing
covariate analysis). First, I did it in the following way (partial
simplified code)

$PRED

THRESH = THETA(1) ;threshold concentration

INT = THETA(2)+ ETA(1)

IF(WCMIN.LE.THRESH) SLOPE = THETA(3)+THETA(4)*CONC+ETA(2)

IF(WCMIN.GT.THRESH) SLOPE = THETA(3)+THETA(5)*CONC+ETA(2)

IPRED=INT+SLOPE*TIME

and NONMEM did not manage to estimate THRESH=THETA(1) (i.e., the initial
estimate did not change during minimization and the gradient was zero
throughout). I tried this for several different initial estimates.

I then centered the observed CONC on THRESH and run the following model
because colleagues mentioned that they have successfully run a
‘hockey-stick’ estimating the threshold parameter  (partial
simplified code)

IF(WCMIN.LE.THRESH) SLOPE = THETA(3)+THETA(4)*(CONC-THRESH)+ETA(2)

IF(WCMIN.GT.THRESH) SLOPE = THETA(3)+THETA(5)*(CONC- THRESH)+ETA(2)

This time, NONMEM

Re: AW: Re: [NMusers] Parameter Estimation in IF Conditioning Statement

[Nick Holford]

Hi Benjamin,

You should not use the closed form  code for a linear model to disease 
progression if the slope is varying with time. See slide 9:

http://holford.fmhs.auckland.ac.nz/docs/clinical-pharmacology-equals-disease-progress-plus-drug-action.pdf

In your case the slope is dependent on WCMIN (which I understand is some 
kind of time varying covariate) and also on CONC (which also seems to be 
time varying.


You need to use an ODE defined model in order to correctly predict the 
time course of the disease when the slope is changing with time. An 
example of NM-TRAN code can be found in slide 39:

http://holford.fmhs.auckland.ac.nz/docs/clinical-pharmacology-equals-disease-progress-plus-drug-action.pdf


If the event records in your NONMEM data set are closely spaced then you 
may get an acceptable approximation because the ADVAN mechanism in 
NONMMEM is equivalent to a very simple piecewise integration.  You can 
also consider using MTIME() in order to perform a more accurate 
piecewise integration.


Best wishes,

Nick

On 20/09/2013 12:42 a.m., Leonid Gibiansky wrote:

Benjamin.
Even as WCMIN is continuous by nature, it is not continuous in the 
dataset (at observation points). You may look at your WCMIN values in 
the data file (I assume this values are in the data file?) and see 
where the initial conditions for THETA(1) placed in that range. Step 
for gradient computations can be quite small, so it will not see 
continuous nature of the variable, and will treat it as discrete.


One the same model but separate topic, I am not sure that you can code 
the model as you coded it if you indeed assume that WCMIN=CONC.

It is better to write as (set aside problems with the gradient):

IF(WCMIN.LE.THRESH) SLOPE = THETA(3)+THETA(4)*CONC+ETA(2)

IF(WCMIN.GT.THRESH) SLOPE = THETA(3)*WCMIN+THETA(5)*(CONC-WCMIN)+ETA(2)

Otherwise, you SLOP function will be discontinuous at WCMIN as a 
function of CONC.


Note also that your two models are not exactly equivalent. If you use
IF(WCMIN.LE.THRESH) SLOPE = THETA(3)+THETA(4)*CONC+ETA(2)
IF(WCMIN.GT.THRESH) SLOPE = THETA(3)+THETA(5)*CONC+ETA(2)

your values at WCMIN=THRESH are equal to
SLOPE = THETA(3)+THETA(4)*THRESH+ETA(2)
SLOPE = THETA(3)+THETA(5)*THRESH+ETA(2)
immediately before and after WCMIN=THRESH (different)

while for the second model they are equal to

SLOPE = THETA(3)+ETA(2)
SLOPE = THETA(3)+ETA(2)

and equal to each other

Leonid


--
Leonid Gibiansky, Ph.D.
President, QuantPharm LLC
web:www.quantpharm.com
e-mail: LGibiansky at quantpharm.com
tel:(301) 767 5566



On 9/19/2013 4:29 PM, benjamin.we...@boehringer-ingelheim.com wrote:
Hi Leonid. Thanks for your quick response. WCMIN is continuous. In 
fact, it is actually identical to CONC (forgot to change it while 
simplifying). Any thoughts on this?Ben


- Originalnachricht -
Von: Leonid Gibiansky [mailto:lgibian...@quantpharm.com]
Gesendet: Thursday, September 19, 2013 09:32 PM
An: Weber,Dr.,Benjamin (TransMed) BIP-DE-B
Cc: nmusers@globomaxnm.com 
Betreff: Re: [NMusers] Parameter Estimation in IF Conditioning Statement

Hi Benjamin
I think the problem was that when nonmem tested small variations of
THETA(1) in the first version, nothing changed in the OF as these
variation have not resulted in the change in the IF conditions (e.g., if
your data contained WCMIN values 1 and 2 and nothing in between, and
THETA(1) was 1.5, changes of THETA(1) to 1.6 or 1.4 were not resulting
in the changes in OF, gradient was zero, and the gradient method was
unable to move the model. When you put theta(1) in the centering, this
resulted in changes of OF. This is not an unusual behavior. Therefore,
it is better to use continuous functions rather than switches. For
example, you can code it as

INT = THETA(4)+(THETA(5)-THETA(4))/(1+(THETA(1)/WCMIN)**GAM)

SLOPE = THETA(3)+INT*CONC+ETA(2)

When GAM is infinity, this is equivalent to your model. You may use
large GAM (or even estimated GAM)

Leonid



--
Leonid Gibiansky, Ph.D.
President, QuantPharm LLC
web:www.quantpharm.com
e-mail: LGibiansky at quantpharm.com
tel:(301) 767 5566



On 9/19/2013 2:31 PM, benjamin.we...@boehringer-ingelheim.com wrote:

Dear NONMEM users,

My goal was to add a drug effect to a disease progression model in a 
way
that the drug effect is different when a certain threshold 
concentration

is exceeded (similarly to modeling a ‘hockey-stick’ when performing
covariate analysis). First, I did it in the following way (partial
simplified code)

$PRED

THRESH = THETA(1) ;threshold concentration

INT = THETA(2)+ ETA(1)

IF(WCMIN.LE.THRESH) SLOPE = THETA(3)+THETA(4)*CONC+ETA(2)

IF(WCMIN.GT.THRESH) SLOPE = THETA(3)+THETA(5)*CONC+ETA(2)

IPRED=INT+SLOPE*TIME

and NONMEM did not manage to estimate THRESH=THETA(1) (i.e., the 
initial

estimate did not change during minimization and the gradient was zero
throughout). I tried this for several different initial est

RE: [NMusers] Parameter Estimation in IF Conditioning Statement

[Mats Karlsson]

Dear Benjamin and Leonid,

Although a steep sigmoidal Emax function often works well to approximate a
step function, the rate of change is quite dependent on the value  of the X
variable. Another function (below) has some advantages. For details, see
Henin et al., AAPS J. 2012 Jun;14(2):155-63. 
A mechanism-based approach for absorption modeling: the Gastro-Intestinal
Transit Time (GITT) model.

STEP = EXP(X*GAM)/(EXP(X*GAM)+EXP(X50*GAM))

Where X50 (inflection point) would be estimated and GAM (steepness)
typically fixed.

Best regards,
Mats
 
Mats Karlsson, PhD
Professor of Pharmacometrics

Dept of Pharmaceutical Biosciences
Faculty of Pharmacy
Uppsala University
Box 591
75124 Uppsala

Phone: +46 18 4714105
Fax + 46 18 4714003
www.farmbio.uu.se/research/researchgroups/pharmacometrics/


-Original Message-
From: owner-nmus...@globomaxnm.com [mailto:owner-nmus...@globomaxnm.com] On
Behalf Of Leonid Gibiansky
Sent: 19 September 2013 21:33
To: benjamin.we...@boehringer-ingelheim.com
Cc: nmusers@globomaxnm.com
Subject: Re: [NMusers] Parameter Estimation in IF Conditioning Statement

Hi Benjamin
I think the problem was that when nonmem tested small variations of
THETA(1) in the first version, nothing changed in the OF as these variation
have not resulted in the change in the IF conditions (e.g., if your data
contained WCMIN values 1 and 2 and nothing in between, and
THETA(1) was 1.5, changes of THETA(1) to 1.6 or 1.4 were not resulting in
the changes in OF, gradient was zero, and the gradient method was unable to
move the model. When you put theta(1) in the centering, this resulted in
changes of OF. This is not an unusual behavior. Therefore, it is better to
use continuous functions rather than switches. For example, you can code it
as

INT = THETA(4)+(THETA(5)-THETA(4))/(1+(THETA(1)/WCMIN)**GAM)

SLOPE = THETA(3)+INT*CONC+ETA(2)

When GAM is infinity, this is equivalent to your model. You may use large
GAM (or even estimated GAM)

Leonid



--
Leonid Gibiansky, Ph.D.
President, QuantPharm LLC
web:www.quantpharm.com
e-mail: LGibiansky at quantpharm.com
tel:(301) 767 5566



On 9/19/2013 2:31 PM, benjamin.we...@boehringer-ingelheim.com wrote:
> Dear NONMEM users,
>
> My goal was to add a drug effect to a disease progression model in a 
> way that the drug effect is different when a certain threshold 
> concentration is exceeded (similarly to modeling a ‘hockey-stick’ when 
> performing covariate analysis). First, I did it in the following way 
> (partial simplified code)
>
> $PRED
>
> THRESH = THETA(1) ;threshold concentration
>
> INT = THETA(2)+ ETA(1)
>
> IF(WCMIN.LE.THRESH) SLOPE = THETA(3)+THETA(4)*CONC+ETA(2)
>
> IF(WCMIN.GT.THRESH) SLOPE = THETA(3)+THETA(5)*CONC+ETA(2)
>
> IPRED=INT+SLOPE*TIME
>
> and NONMEM did not manage to estimate THRESH=THETA(1) (i.e., the 
> initial estimate did not change during minimization and the gradient 
> was zero throughout). I tried this for several different initial
estimates.
>
> I then centered the observed CONC on THRESH and run the following 
> model because colleagues mentioned that they have successfully run a 
> ‘hockey-stick’ estimating the threshold parameter  (partial simplified 
> code)
>
> IF(WCMIN.LE.THRESH) SLOPE = THETA(3)+THETA(4)*(CONC-THRESH)+ETA(2)
>
> IF(WCMIN.GT.THRESH) SLOPE = THETA(3)+THETA(5)*(CONC- THRESH)+ETA(2)
>
> This time, NONMEM provided very reasonable estimates for all model 
> parameters and the covariance step was successful.
>
> I wonder now if the centering of the variable (and hence something 
> particular to the data set) caused the difference in estimability of 
> the threshold parameter or whether NONMEN cannot estimate parameters 
> that only occur in the conditioning part of the IF statement (Note 
> that after centering THRESH also appears in another part of the code).
>
> Could somebody please provide some insight on this? I have searched in 
> the NONMEM user group but could not find anything.
>
> Mit freundlichen Grüßen / Kind regards, Dr. Benjamin Weber
>
> Boehringer Ingelheim Pharma GmbH & Co. KG Translational Medicine
> Tel.: +49 (7351) 54-143520
> Fax: +49 (7351) 83-143520
> mailto:benjamin.we...@boehringer-ingelheim.com
>
> Boehringer Ingelheim Pharma GmbH & Co. KG, Sitz: Ingelheim am Rhein; 
> Registergericht Mainz: HR A 22206; Komplementär Boehringer Ingelheim 
> Deutschland GmbH; Geschäftsführung: Dr. Engelbert Günster 
> (Vorsitzender),  Ursula Fuggis-Hahn, Ralf Gorniak,  Michael Klein, Dr.
> Martin Wanning; Vorsitzender des Aufsichtsrates: Dr. Joachim 
> Hasenmaier;
> Sitz: Ingelheim am Rhein; Registergericht Mainz: HR B 23260
>
> Diese E-Mail ist vertraulich zu behandeln. Sie kann besonderem 
> rechtlichem Schutz unterliegen. Wenn Sie nicht der richtige Adressat 
> sind, senden Sie bitte diese E-Mail an den Absender zurück, löschen 
> die eingegangene E-Mail und geben den Inhalt der E-Mail nicht weiter.
> Jegliche unbefugte Bearbeitung, Nutzung, Vervielfältigung o