Hi everyone,
[D]t={(F0-F/F)*(1/Kb)}^(1/n)+n*[P]t*(F0-F/F)
where [D]t and [P]t gives total drug conc and total protein conc. we want to determine Kb and n. we have a series of data for y=(F0-F/F), x=[D]t (our two variables).how can one fit this equation? Is there any other paper where this equation has been used?
Plz help me out if u know.
You can fit the data in Origin with user defined function
https://www.originlab.com/doc/Tutorials/UserDef-FitFunc
Alternatively you can try to rearrange the equation to get a linear form
[D]t={(F0-F/F)*(1/Kb)}^(1/n)+n*[P]t*(F0-F/F)
Taking (F0-F/F)^(1/n) common
[D]t=(F0-F/F)^(1/n){(1/Kb)}^(1/n)+n*[P]t*(F0-F/F)^(1-1/n)}
Taking log
log[D]t=1/n*log(F0-F/F) + log{(1/Kb)}^(1/n)+n*[P]t*(F0-F/F)^(1-1/n)}
Plot of log(F0-F/F) vs log[D]t will give the slope of "1/n" and intercept of "log{(1/Kb)}^(1/n)+n*[P]t*(F0-F/F)^(1-1/n)}". In the intercept the unknowns are "n" (already found from slope) and "Kb". Only problem here is which value of "(F0-F/F)" to use. Please check yourself whether the rearrangement is done properly on not (as I just did it myself).
I think you can easily use Origin to fit nonlinear data of two unknowns.
Thank u Tuhin Khan . I have already used origin6.0 by user defined eqn. but at the end they ask to put the value of Kb and n before iteration start. what value should i put there as it is already a unknown parameter we want to determine?
according to me the log eqn u tried is not right probably. I have also tried to fit this by taking log. it did not work.
Thanks once again.
I see the problem in my formulation. The intercept is not a constant it contains the variable (F0-F/F). So it will not work. Then I think direct nonlinear fit is the only option.
Of course you have to give a initial guess value. The guess value do not have to be correct but it has to be in same order of your expected result. You can the take the initial value from literature. I think there must be lot of such values in literature and when you did this experiment you are probably expecting a certain result.
If "n" is some kind of association number then give a initial value of 1-5 and if "Kb" is binding constant take the value from literature (it is not possible for me to make a guess for your system).
As you say i had put the initial value of n and Kb but it shows the error "poor fitting parameter something like that".
Anyway i will try in another way. Lets see .
Thank u for ur response!
- Badges
- Science topic
- Similar topics
- Chemistry
- Pharmacology
- Pharmaceuticals
- Drugs