Jump to content


Photo

Secondary parameters


  • Please log in to reply
7 replies to this topic

#1 raghava choudary

raghava choudary

    Advanced Member

  • Members
  • PipPipPip
  • 81 posts

Posted 29 August 2015 - 04:33 AM

Hi all,

 

How to generate secondary parameters for a 2 compartment model using phoenix Model??

 

Thanks in advance...

 

Raghav

 



#2 Simon Davis

Simon Davis

    Advanced Member

  • Administrators
  • 1,316 posts

Posted 29 August 2015 - 10:02 AM

Hi Raghav, a simple way to get started when using standard library/built in models is to the use the set_WNL_model button and you will notice this populates the secondary parameters.

 

If you need more, or you are creating a new model this is where you add the formula for that secondary parameter.

 

   Simon

Attached Thumbnails

  • set_wnl_model.jpg
  • seco_model12.jpg


#3 raghava choudary

raghava choudary

    Advanced Member

  • Members
  • PipPipPip
  • 81 posts

Posted 29 August 2015 - 12:59 PM

Attached File  PK Modeling 290815.phxproj   447.98KB   376 downloads

 

Hi Simon,

 

Thank you for the reply,

 

Used the Set_WNL Model and a set of secondary parameters were generated. However some parameters that are related to 3 comp were also added. The standard errors and % CV are high. 

 

In theta tab for some estimates the 95 % CI includes negative values ???

 

Regards,

 

Raghav



#4 Simon Davis

Simon Davis

    Advanced Member

  • Administrators
  • 1,316 posts

Posted 29 August 2015 - 09:01 PM

Raghav, when changing models I would generally recommend choosing 'replace' so that you don't 'keep' parameters which are no longer relevant to the new model as you seem to have done in this case.

 

otherwise I might be tempted to consider some modifcation to the model e.g. zero-order absorption?

 

 Simon.



#5 Simon Davis

Simon Davis

    Advanced Member

  • Administrators
  • 1,316 posts

Posted 02 September 2015 - 12:33 PM

Raghav,

       Another option if you want to avoid negative values is to transform your parameter in the log domain.  For example

Cllog =log(Cl)
In the model you will have Cl replaced by exp(Cllog)

 

You will get 'raw' confidence intervals for Cllog that can be negative then exponentiate it and they will be positive.

 

  Simon.



#6 raghava choudary

raghava choudary

    Advanced Member

  • Members
  • PipPipPip
  • 81 posts

Posted 03 September 2015 - 04:38 AM

Raghav,

       Another option if you want to avoid negative values is to transform your parameter in the log domain.  For example

Cllog =log(Cl)
In the model you will have Cl replaced by exp(Cllog)

 

You will get 'raw' confidence intervals for Cllog that can be negative then exponentiate it and they will be positive.

 

  Si

 

Raghav, when changing models I would generally recommend choosing 'replace' so that you don't 'keep' parameters which are no longer relevant to the new model as you seem to have done in this case.

 

otherwise I might be tempted to consider some modifcation to the model e.g. zero-order absorption?

 

 Simon.

Hi Simon,

 

Considered 2 compartment , Zero order absorption  model and still see that the precision on parameters V2 and CL2 is very low (high % CV) !!

 

Am i missing anything , plz find the attached model. Attached File  PK Modeling 290815.phxproj   1.07MB   374 downloads

 

How can i get secondary parameters when Editing Graphically the model.

 

thanks in advance ,

 

Raghav



#7 Simon Davis

Simon Davis

    Advanced Member

  • Administrators
  • 1,316 posts

Posted 03 September 2015 - 10:23 AM

Hi Raghav, secondary parameters can be added in Graphical model the same way as in Built-in/library model mode like  before.  Remember if starting from a built-in model you'll need to check that edits you've made to the model don't make the calculation formulas invalid for any existing secondary parameters.

 

Regarding the higher CV% of V2 etc, this may just be an identifiability issue as you are not directly observing this compartment but I'll ask a colleague if he has any other suggestions.

 

  Simon.

Attached Thumbnails

  • seco.jpg


#8 raghava choudary

raghava choudary

    Advanced Member

  • Members
  • PipPipPip
  • 81 posts

Posted 05 September 2015 - 10:39 AM

Hi Raghav, secondary parameters can be added in Graphical model the same way as in Built-in/library model mode like  before.  Remember if starting from a built-in model you'll need to check that edits you've made to the model don't make the calculation formulas invalid for any existing secondary parameters.

 

Regarding the higher CV% of V2 etc, this may just be an identifiability issue as you are not directly observing this compartment but I'll ask a colleague if he has any other suggestions.

 

  Simon.

 

Raghav, when changing models I would generally recommend choosing 'replace' so that you don't 'keep' parameters which are no longer relevant to the new model as you seem to have done in this case.

 

otherwise I might be tempted to consider some modifcation to the model e.g. zero-order absorption?

 

 Simon.

 

Hi All,

 

I have a general question on Zero order absorption. If a compound follows zero order absorption and has lag time (say for e.g., 1 hr), will the tmax change upon increasing the doses ?? Will Cmax increase proportionately to initial dose ??

How they are comparable with respect to first order absorption ?

 

Best regards,

 

Raghav,






0 user(s) are reading this topic

0 members, 0 guests, 0 anonymous users