Auto-suggest helps you quickly narrow down your search results by suggesting possible matches as you type.

- Subscribe to RSS Feed
- Mark Topic as New
- Mark Topic as Read
- Float this Topic for Current User
- Bookmark
- Subscribe
- Printer Friendly Page

Highlighted
##

- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Get Direct Link
- Email to a Friend
- Report Inappropriate Content

DOE - ELISA method - 4 PL modeling

Nov 27, 2018 7:46 AM
(3774 views)

Dear JMP Community,

My question is about a DOE performed on ELISA method.

The response is the Absorbance and there are 3 factors which are dilution of different reagents (standard, parameter B and parameter C) - Refer to graph attached to visualize the type of data I have.

I’m looking for optimizing the three factors to get a plateau equal to 2 for absorbance. But I would like also to be at “the beginning” of the plateau.

I managed to optimize to get a plateau around 2 (I used 4PL modeling). But I don’t know how to be at the beginning of the plateau. Would you have ideas to share to parameter this need?

Thanks

5 REPLIES 5

Highlighted
##

- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Get Direct Link
- Email to a Friend
- Report Inappropriate Content

Re: DOE - ELISA method - 4 PL modeling

Not exactly sure what you mean by "be at the beginning of the plateau". As I understand, you have found conditions where Absorbance vs log(1/[standard]) plateaus at 2.

Do you mean that you would like the plateau to begin at a lower value of log(1/[standard]) ?

Do you mean that you would like the plateau to begin at a lower value of log(1/[standard]) ?

Highlighted
##

- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Get Direct Link
- Email to a Friend
- Report Inappropriate Content

Re: DOE - ELISA method - 4 PL modeling

Hi,

I updated the document attached with a new schema to detail what I mean by "beginning of the plateau".

It could correspond to an IC99 for example.

Is it clearer for you?

Highlighted
##

- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Get Direct Link
- Email to a Friend
- Report Inappropriate Content

Re: DOE - ELISA method - 4 PL modeling

Hi.

That helps but I am still not sure what your objective is. It is clear what you mean by "the beginning of the plateau". But I am still not clear what you mean by "be at the beginning of the plateau", with "be at" being the bit I don't get.

Phil

That helps but I am still not sure what your objective is. It is clear what you mean by "the beginning of the plateau". But I am still not clear what you mean by "be at the beginning of the plateau", with "be at" being the bit I don't get.

Phil

Highlighted
##

- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Get Direct Link
- Email to a Friend
- Report Inappropriate Content

Re: DOE - ELISA method - 4 PL modeling

I'll try to be clearer :

- The response studied is the absorbance.

It is modeled by a 4PL curve when represented in function of the dilution of the standard used in the analysis: Absorbance = f (Log 1/[standard])

- I'm analysing three factors:

- Dilution of the standard (x-axis)

- Parameter B (dilution of a reagent used in the analysis)

- Parameter C (dilution of a second reagent used in the analysis)

As we can observe on the first schema, Parameters B and C will mostly affect the value of the plateau (or also difference min-max of the curves).

The ideal configuration I'm looking for is the point that check 2 conditions

- Absorbance plateau = 2

- Dilution of the standard corresponding to the beginning of plateau (IC99 maybe)

At the end, I would like to define value of the three factors that allows me to reach the ideal configuration.

My issue is that I don't know how to configure this ideal configuration in the profiler/desirability to get this optimization.

Highlighted
##

- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Get Direct Link
- Email to a Friend
- Report Inappropriate Content

Re: DOE - ELISA method - 4 PL modeling

That is completely clear now, thanks.

For simplicity let's call your factors: [std], [rgt1], [rgt2]

You had said that you had found settings of [rgt1] and [rgt2] such that absorbance vs [std] plateaus at 2.

So now all you need to do is find the setting of [std] that corresponds to IC99, that is 99% of 2 = 1.98. How did you fit the 4PL curves? What platform in JMP did you use? Depending on how you fit the model it should be easy to find the IC99 using either the profiler or inverse prediction.

For simplicity let's call your factors: [std], [rgt1], [rgt2]

You had said that you had found settings of [rgt1] and [rgt2] such that absorbance vs [std] plateaus at 2.

So now all you need to do is find the setting of [std] that corresponds to IC99, that is 99% of 2 = 1.98. How did you fit the 4PL curves? What platform in JMP did you use? Depending on how you fit the model it should be easy to find the IC99 using either the profiler or inverse prediction.

Article Labels

There are no labels assigned to this post.