Re: [NMusers] backward integration from t-a to t

2014-01-17 Thread Nick Holford 

Pavel,

It seems you prefer empirical curve fitting to science based modelling 
because you do not seem to think that my suggestions were constructive. 
I am glad I don't have your job.


Nick

On 18/01/2014 5:30 a.m., Pavel Belo wrote:

Thank you Nick.
We see the advantage and disadvantages this approach clear and 
understand the difference between the modeling and the curve fitting 
exercise.  On the other hand, out motto is to keep an open mind and 
to  keep trying.  There are scenarios where the approach may work and 
where it will not work.  In any case, it is useful to study it and 
keep it in the library even if it is classified as a rare case.
It is easier to provide critique than to suggest something 
constructive.  Lets phrase Leonid for doing the constructive part and 
being innovative.  Lets thank Robert for providing the algorithm!

Take care,
Pavel
On Thu, Jan 16, 2014 at 07:48 PM, Nick Holford wrote:

Pavel,
Unless your drug is an alkylating agent the use of AUC will always
be mechanistically wrong.
I hope you also considered the possibility of disease progression
(i.e. changing baseline) and also the possibility of changing C50
due to potentiation or physiological changes.
Nick

On 17/01/2014 1:29 p.m., lgibian...@quantpharm.com wrote:

Hi Pavel,
You mentioned that the effect compartment did not help, and the model I
suggested is identical to the effect compartment. May be try something like
transit compartment model:

DADT(2)=C-K0*A(2)
DADT(3)=K0*A(2)-K0*A(3)
...
DADT(X)=K0*A(X-1)-K0*A(X)

AUCapprox=A(2)+...+A(X)

This will prolong the shape of AUCapprox(t). It could be a bit simpler and
smoother than tlag implementation
Leonid







Original email:
-
From: Pavel belonon...@optonline.net
Date: Thu, 16 Jan 2014 13:05:54 -0500 (EST)
To:lgibian...@quantpharm.com,nmusers@globomaxnm.com
Subject: RE: [NMusers] backward integration from t-a to t


Hello Leonid,

Thank you bein helpful.  You got the main point.  AUC is a better
predictor than concentration, but it has to disppear very slowly but
surely.

A potential challenge is biological meaning of this approach.  It will
be necessary to explain it to the biologists, who ask question like "Why
do you use 2 compartment in PK model while human body has so many
compartments?".

We will see!

Thanks,
Pavel



On Wed, Jan 15, 2014 at 01:19 AM,lgibian...@quantpharm.com  wrote:


Pavel,
I think one can use equation
DADT(2)=C-K0*A(2)

where C is the drug concentration. When K0=0, A2 is cumulative AUC.
When
k0>0, A2 would represent something like AUC for the interval prior to
the current
time
The length of the interval would be proportional to 1/K0 (and equal to
infinity when k0=0). Conceptually, K0 is the rate of "AUC elimination"
from the
system. PD then can be made dependent on A2, and the model would
select optimal
value of K0. One interesting case to understand the concept is when C
is constant.
Then A2=C/K0 while AUC over some interval TAU is AUC=C*TAU. So
roughly, A2 can
be interpreted as AUC over the interval of 1/K0. Leonid


Original email:
-
From: Pavel belonon...@optonline.net
Date: Tue, 14 Jan 2014 13:45:18 -0500 (EST)
To:robert.ba...@iconplc.com,nmusers@globomaxnm.com
Subject: [NMusers] backward integration from t-a to t





Dear Robert,




Ã,

Efficacy isÃ, frequently considered aÃ, function of AUC.Ã, (AUC is just
an integral. It is obvious how to calculate AUC any software which can
solve ODE.)Ã, A disadvantage of this model of efficacyÃ, is that the
effect is irreversable becauseÃ, AUC of concentration can only
increase;Ã, it cannot decrease.Ã, In many cases, a more meaningful model
is a model where AUC is calculated form time tÃ, -a to t (kind of
"moving average"), where t is timeÃ, in the system of differential
equations (variable T in NONMEM).Ã, Ã, There are 2 obvious ways to
calculate AUC(t-a, t).Ã, The first is to do backward integration, which
looks like a hard and resource consuming way for NONMEM.Ã, The second
one is to keep in memory AUC for all time pointsÃ, usedÃ, during theÃ,
integrationÃ, and calculate AUC(t-a,t) as AUC(t) - AUC(t-a), there
AUC(t-a) can be interpolated using two closest time points below and
above t-a.Ã,

Ã,

Is there a way toÃ, access AUC forÃ, the past time points (> integration
routine?Ã, It seems like an easyÃ, thing to do.Ã, Ã, Ã,

Ã,

Kind regards,


PavelÃ, Ã,



mail2web - Check your email from the web at
http://link.mail2web.com/mail2web





myhosting.com - Premium Microsoft® Wind

[NMusers] ACoP5 Call for Programming and Workshop Proposals

2014-01-17 Thread ISoP Announcements 
The call for workshop proposals is now open!  The ACoP5 Planning Committee
invites proposals from parties who are interested in conducting educational
workshops on topics related to pharmacometrics, to be held immediately
before and after ACoP5 in Las Vegas, October 12-15th, 2014.  We encourage
you to contact the ACoP5 Planning Committee with ideas for workshops that
will benefit the pharmacometric community, and get more information on
submitting a workshop proposal by visiting the website at
http://www.acop5.org/workshops.

 

We would also like to remind you that the call for programming for the ACoP5
meeting October 12-15, 2014 is now open and the programming committee is
accepting proposals.  The meeting organizers are encouraging proposals
relating to practical, methodological and theoretical aspects of
pharmacometrics in a symposia, round table or tutorial format.  Please visit
the call for programming website at
http://www.acop5.org/call-for-programming for more details.  

 

If you have any questions please do not hesitate to contact the programming
committee at acop5programm...@go-isop.org.

Re: [NMusers] Taca Training Workshop

2014-01-17 Thread Paul Hutson 

Users:
FWIW, this is an excellent workshop. I took it last year, and wish I had 
been able to take something like it decades ago.

Venue is convenient to Dublin, too.
Paul

On 1/17/2014 4:55 AM, Adrian Dunne wrote:


   TACA TRAINING www.tacatraining.com

 PHARMACOMETRIC STATISTICS

Registration is now open for this 3 day workshop to be held from
  8th to 10th October 2014 in Dublin, Ireland.

The aim of this 3 day workshop is to give pharmacometricians a good
understanding of the statistical concepts upon which their work is
based and which are of great importance in everything they do. The
emphasis will be on concepts with an absolute minimum of mathematical
details. Attendees need only have studied statistics at foundation
level prior to taking this course. The topics covered include;

1) Why use statistics?

2) Probability and statistical inference.

3) Laws of probability and Bayes theorem.

4) Univariate probability distributions – Expected value and variance.

5) Multivariate probability distributions – joint, marginal and
conditional distributions. The covariance matrix. Independence and
conditional independence.

6) Modelling, estimation, estimators, sampling distributions, bias,
efficiency, standard error and mean squared error. Consistency.

7) Point and interval estimators. Confidence intervals.

8) Hypothesis testing, null and alternative hypotheses. P-value, Type
I and type II errors and power.

9) Likelihood inference, maximum likelihood estimator (MLE),
likelihood ratio. BQL and censored data.

10) Minimal sufficiency and invariance of the likelihood ratio and the MLE.

11) The score function, hessian, Fisher information, quadratic
approximation and standard error.

12) Wald confidence intervals and hypothesis tests.

13) Likelihood ratio tests.

14) Profile likelihood, nested models.

15) Model selection, Akaike and Bayesian Information Criteria (AIC & BIC).

16) Maximising the likelihood, Newton’s method.

17) Mixed effects models.

18) Estimation of the fixed effects, conditional independence, prior
and posterior distributions.

19) Approximating the integrals, Laplace and first order (FO & FOCE)
approximations, numerical quadrature.

20) The Expectation Maximisation (EM) algorithm.

21) Estimating the random effects, empirical bayes estimates (EBE) and
shrinkage.

22) Asymptotic properties of the MLE, efficiency, the Cramer-Rao Lower
Bound (CRLB), consistency, normality.

23) Robustness of the MLE, the Kullback-Liebler distance. Quasi
likelihood and the robust or sandwich variance estimator.

24) Time to event (survival) analysis. Survivor and Hazard functions.

25) Kaplan-Meier plots. Log-rank and Wilcoxon tests.

26) Parametric and semi-parametric proportional hazards models.

27) Partial and full likelihood inference.

For further details and to register please go to our website
www.tacatraining.com

Early registration is advised because the number of places is limited.

Adrian Dunne

**
  Taca Training

   Advanced Training Workshops
   for the Pharmaceutial Industry

6 The Avenue, Woodpark, Ballinteer, Dublin 16, Ireland.
Phone +353-(0)1-2986843
Mobile +353-(0)86-0407504

www.tacatraining.com
 adrian.du...@tacatraining.com

**





--
Paul R. Hutson, Pharm.D.
Professor
Senior Associate Dean for Academic Affairs
UW School of Pharmacy
T: 608.263.2496
F: 608.265.5421

RE: [NMusers] backward integration from t-a to t

2014-01-17 Thread Pavel Belo 

Hello Leonid,

I agree that the effect compartment model somewhat resembles the ALAG 
approach.  Sometimes they can provide very similar outcome.  We have the 
resolution and I'll update you when it is appropriate.  Unfortunately, I 
cannot reveal some information right now.  Thanks!


Kind regards,
Pavel


On Thu, Jan 16, 2014 at 07:29 PM, lgibian...@quantpharm.com wrote:


Hi Pavel,
You mentioned that the effect compartment did not help, and the model 
I suggested is identical to the effect compartment. May be try 
something like transit compartment model:


DADT(2)=C-K0*A(2)
DADT(3)=K0*A(2)-K0*A(3)
...
DADT(X)=K0*A(X-1)-K0*A(X)

AUCapprox=A(2)+...+A(X)

This will prolong the shape of AUCapprox(t). It could be a bit simpler 
and smoother than tlag implementation Leonid







Original email:
-
From: Pavel Belo non...@optonline.net
Date: Thu, 16 Jan 2014 13:05:54 -0500 (EST)
To: lgibian...@quantpharm.com, nmusers@globomaxnm.com
Subject: RE: [NMusers] backward integration from t-a to t


Hello Leonid,

Thank you bein helpful. You got the main point. AUC is a better 
predictor than concentration, but it has to disppear very slowly but 
surely.


A potential challenge is biological meaning of this approach. It will 
be necessary to explain it to the biologists, who ask question like 
"Why do you use 2 compartment in PK model while human body has so many 
compartments?".


We will see!

Thanks,
Pavel



On Wed, Jan 15, 2014 at 01:19 AM, lgibian...@quantpharm.com wrote:


Pavel,
I think one can use equation
DADT(2)=C-K0*A(2)

where C is the drug concentration. When K0=0, A2 is cumulative AUC. 
When
k0>0, A2 would represent something like AUC for the interval prior to 
the current

time
The length of the interval would be proportional to 1/K0 (and equal 
to
infinity when k0=0). Conceptually, K0 is the rate of "AUC 
elimination" from the
system. PD then can be made dependent on A2, and the model would 
select optimal
value of K0. One interesting case to understand the concept is when C 
is constant.
Then A2=C/K0 while AUC over some interval TAU is AUC=C*TAU. So 
roughly, A2 can

be interpreted as AUC over the interval of 1/K0. Leonid


Original email:
-
From: Pavel Belo non...@optonline.net
Date: Tue, 14 Jan 2014 13:45:18 -0500 (EST)
To: robert.ba...@iconplc.com, nmusers@globomaxnm.com
Subject: [NMusers] backward integration from t-a to t





Dear Robert,




Â

Efficacy is frequently considered a function of AUC. (AUC is 
just an integral. It is obvious how to calculate AUC any software 
which can solve ODE.) A disadvantage of this model of efficacy is 
that the effect is irreversable because AUC of concentration can 
only increase; it cannot decrease. In many cases, a more 
meaningful model is a model where AUC is calculated form time t -a 
to t (kind of "moving average"), where t is time in the system of 
differential equations (variable T in NONMEM).  There are 2 
obvious ways to calculate AUC(t-a, t). The first is to do backward 
integration, which looks like a hard and resource consuming way for 
NONMEM. The second one is to keep in memory AUC for all time 
points used during the integration and calculate AUC(t-a,t) 
as AUC(t) - AUC(t-a), there AUC(t-a) can be interpolated using two 
closest time points below and above t-a.Â


Â

Is there a way to access AUC for the past time points (> 
integration routine? It seems like an easy thing to do.  Â


Â

Kind regards,


Pavel Â



mail2web - Check your email from the web at
http://link.mail2web.com/mail2web






myhosting.com - Premium Microsoft® Windows® and Linux web and 
application

hosting - http://link.myhosting.com/myhosting

Re: [NMusers] backward integration from t-a to t

2014-01-17 Thread Pavel Belo 




Thank you Nick. 


 


We see the advantage and disadvantages this approach clear and 
understand the difference between the modeling and the curve fitting 
exercise.  On the other hand, out motto is to keep an open mind and to  
keep trying.  There are scenarios where the approach may work and where 
it will not work.  In any case, it is useful to study it and keep it in 
the library even if it is classified as a rare case. 



 


It is easier to provide critique than to suggest something 
constructive.  Lets phrase Leonid for doing the constructive part and 
being innovative.  Lets thank Robert for providing the algorithm! 



 


Take care,


Pavel


 


 
On Thu, Jan 16, 2014 at 07:48 PM, Nick Holford wrote:

 


 

Pavel,
Unless your drug is an alkylating agent the use of AUC will always be 
mechanistically wrong.
I hope you also considered the possibility of disease progression (i.e. 
changing baseline) and also the possibility of changing C50 due to 
potentiation or physiological changes.

Nick


On 17/01/2014 1:29 p.m., lgibian...@quantpharm.com 
  wrote:


Hi Pavel,
You mentioned that the effect compartment did not help, and the model I
suggested is identical to the effect compartment. May be try something 
like

transit compartment model:

DADT(2)=C-K0*A(2)
DADT(3)=K0*A(2)-K0*A(3)
...
DADT(X)=K0*A(X-1)-K0*A(X)

AUCapprox=A(2)+...+A(X)

This will prolong the shape of AUCapprox(t). It could be a bit simpler 
and

smoother than tlag implementation
Leonid







Original email:
-
From: Pavel Belo non...@optonline.net 
Date: Thu, 16 Jan 2014 13:05:54 -0500 (EST)
To: lgibian...@quantpharm.com  , 
nmusers@globomaxnm.com 

Subject: RE: [NMusers] backward integration from t-a to t


Hello Leonid,

Thank you bein helpful.  You got the main point.  AUC is a better
predictor than concentration, but it has to disppear very slowly but
surely.

A potential challenge is biological meaning of this approach.  It will
be necessary to explain it to the biologists, who ask question like "Why
do you use 2 compartment in PK model while human body has so many
compartments?".

We will see!

Thanks,
Pavel



On Wed, Jan 15, 2014 at 01:19 AM, lgibian...@quantpharm.com 
  wrote:



Pavel,
I think one can use equation
DADT(2)=C-K0*A(2)

where C is the drug concentration. When K0=0, A2 is cumulative AUC.
When
k0>0, A2 would represent something like AUC for the interval prior to
the current
time
The length of the interval would be proportional to 1/K0 (and equal to
infinity when k0=0). Conceptually, K0 is the rate of "AUC elimination"
from the
system. PD then can be made dependent on A2, and the model would
select optimal
value of K0. One interesting case to understand the concept is when C
is constant.
Then A2=C/K0 while AUC over some interval TAU is AUC=C*TAU. So
roughly, A2 can
be interpreted as AUC over the interval of 1/K0. Leonid


Original email:
-
From: Pavel Belo non...@optonline.net 
Date: Tue, 14 Jan 2014 13:45:18 -0500 (EST)
To: robert.ba...@iconplc.com  , 
nmusers@globomaxnm.com 

Subject: [NMusers] backward integration from t-a to t





Dear Robert,




Ã,

Efficacy isÃ, frequently considered aÃ, function of AUC.Ã, (AUC is just
an integral. It is obvious how to calculate AUC any software which can
solve ODE.)Ã, A disadvantage of this model of efficacyÃ, is that the
effect is irreversable becauseÃ, AUC of concentration can only
increase;Ã, it cannot decrease.Ã, In many cases, a more meaningful model
is a model where AUC is calculated form time tÃ, -a to t (kind of
"moving average"), where t is timeÃ, in the system of differential
equations (variable T in NONMEM).Ã, Ã, There are 2 obvious ways to
calculate AUC(t-a, t).Ã, The first is to do backward integration, which
looks like a hard and resource consuming way for NONMEM.Ã, The second
one is to keep in memory AUC for all time pointsÃ, usedÃ, during theÃ,
integrationÃ, and calculate AUC(t-a,t) as AUC(t) - AUC(t-a), there
AUC(t-a) can be interpolated using two closest time points below and
above t-a.Ã,

Ã,

Is there a way toÃ, access AUC forÃ, the past time points (> integration
routine?Ã, It seems like an easyÃ, thing to do.Ã, Ã, Ã,

Ã,

Kind regards,


PavelÃ, Ã,



mail2web - Check your email from the web at
http://link.mail2web.com/mail2web 




myhosting.com - Premium Microsoft® Windows® and Linux web and 
application
hosting - http://link.myhosting.com/myhosting 





--
Nick Holford, Professor Clinical Pharmacology
Dept Pharmacology & Clinical Pharmacology, Bldg 503 R

[NMusers] Taca Training Workshop

2014-01-17 Thread Adrian Dunne 


  TACA TRAINING www.tacatraining.com

PHARMACOMETRIC STATISTICS

Registration is now open for this 3 day workshop to be held from
 8th to 10th October 2014 in Dublin, Ireland.

The aim of this 3 day workshop is to give pharmacometricians a good
understanding of the statistical concepts upon which their work is
based and which are of great importance in everything they do. The
emphasis will be on concepts with an absolute minimum of mathematical
details. Attendees need only have studied statistics at foundation
level prior to taking this course. The topics covered include;

1) Why use statistics?

2) Probability and statistical inference.

3) Laws of probability and Bayes theorem.

4) Univariate probability distributions – Expected value and variance.

5) Multivariate probability distributions – joint, marginal and
conditional distributions. The covariance matrix. Independence and
conditional independence.

6) Modelling, estimation, estimators, sampling distributions, bias,
efficiency, standard error and mean squared error. Consistency.

7) Point and interval estimators. Confidence intervals.

8) Hypothesis testing, null and alternative hypotheses. P-value, Type
I and type II errors and power.

9) Likelihood inference, maximum likelihood estimator (MLE),
likelihood ratio. BQL and censored data.

10) Minimal sufficiency and invariance of the likelihood ratio and the MLE.

11) The score function, hessian, Fisher information, quadratic
approximation and standard error.

12) Wald confidence intervals and hypothesis tests.

13) Likelihood ratio tests.

14) Profile likelihood, nested models.

15) Model selection, Akaike and Bayesian Information Criteria (AIC & BIC).

16) Maximising the likelihood, Newton’s method.

17) Mixed effects models.

18) Estimation of the fixed effects, conditional independence, prior
and posterior distributions.

19) Approximating the integrals, Laplace and first order (FO & FOCE)
approximations, numerical quadrature.

20) The Expectation Maximisation (EM) algorithm.

21) Estimating the random effects, empirical bayes estimates (EBE) and
shrinkage.

22) Asymptotic properties of the MLE, efficiency, the Cramer-Rao Lower
Bound (CRLB), consistency, normality.

23) Robustness of the MLE, the Kullback-Liebler distance. Quasi
likelihood and the robust or sandwich variance estimator.

24) Time to event (survival) analysis. Survivor and Hazard functions.

25) Kaplan-Meier plots. Log-rank and Wilcoxon tests.

26) Parametric and semi-parametric proportional hazards models.

27) Partial and full likelihood inference.

For further details and to register please go to our website
www.tacatraining.com

Early registration is advised because the number of places is limited.

Adrian Dunne

**
 Taca Training

  Advanced Training Workshops
  for the Pharmaceutial Industry

6 The Avenue, Woodpark, Ballinteer, Dublin 16, Ireland.
Phone +353-(0)1-2986843
Mobile +353-(0)86-0407504

   www.tacatraining.com
adrian.du...@tacatraining.com

**