I carry out ELISA experiments, in which a log/log curve fit is required in order to calculate concentration from O.D.
I have used the drc package previously to fit a 4PL model, but this kit requires a loglog curve fit.
I am not sure if I am on the correct track or if my formula is even correct. The table is not correct, most of the values are NA, and that is not what I expect.
This is my code so far:
od
Hi Mona. Check this site. It may help you.
https://katatrepsis.com/2018/09/24/elisa-analysis-in-r/
Thank you Kerim Goztepe unfortunetly this link discusses the 4PL regression model, when I am looking to do a log/log curve.
Thank you very much Jochen Wilhelm , how can I predict the concentrations from the list of ODs I have? I am not quite sure by the scatterplot with the data part, could you kindly clarify?
You implied to express the measured OD by the given concentration. If you have the OD given and you want to know the concentration, you would have to find the inverse function. There might be some smart trick in drc, but I solved the problem "manually" with a function that searches the x-values (conc) at which a given y-value (od) is obtained:
inverse
Hi Marierose,
Try this, it will generate a graph and fit parameters for a 4 parameter curve fit, no need for log transformation. It can also calculate unknown values. You need ot install the packages http, ggplot2 and rhandsontable (if you want results in a nice spreadsheet)
You mention you need a log/log curve fit, does the kit specify it is plotted as log/log or is the fit function some kind of a log function?
library(httr)
library(ggplot2)
library(rhandsontable)
od
Jochen Wilhelm I can't neither find an implementation of the inverse function in drc package. It seems odd to me why they didn't provide it.
The practice of fitting conc~od is really wrong but I can see many people doing this.
- Badges
- Science method
- Similar topics
- Biological Science
- Immunology
- Immunology Techniques
- Immunoassay